%/cs/research/crest/home1/ucacbbl/bibtex/IEEEabrv.bib processed by add_key_xplor.awk Revision: 1.3 $ 25 Aug 2021
%/tmp/cec2021.bib4 processed by add_key_xplor.awk Revision: 1.3 $ 25 Aug 2021
%$0 $Revision: 1.12 $  eden.cs.ucl.ac.uk /cs/research/crest/home1/ucacbbl/doc/refs Wed Aug 25 08:19:30 BST 2021


@INPROCEEDINGS(:2021:CEC,
%xplor 25 Aug 2021
  author = {},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={[{CEC} 2021 Front cover]},
year={2021},
  editor = {Yew-Soon Ong},
pages = {c1--c1},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Presents the front cover or splash screen of the proceedings
record.},
keywords = {},
doi = {doi:10.1109/CEC45853.2021.9505005},
  notes = {Also known as \cite{9505005}},
)


@INPROCEEDINGS(:2021:CEC,
%xplor 25 Aug 2021
  author = {},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={[{CEC} 2021 copyright notice]},
year={2021},
  editor = {Yew-Soon Ong},
pages = {c1--c1},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Presents the copyright information for the conference. May
include reprint permission information.},
keywords = {},
doi = {doi:10.1109/CEC45853.2021.9504727},
  notes = {Also known as \cite{9504727}},
)


@INPROCEEDINGS(:2021:CEC,
%xplor 25 Aug 2021
  author = {},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Welcome Message},
year={2021},
  editor = {Yew-Soon Ong},
pages = {i--i},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The 2021 IEEE Congress on Evolutionary Computation (IEEE CEC
2021) was planned to be held in Krakow, Poland, however, due to COVID-19
restrictions, it is held virtually, between June 28th and July 1st 2021.
IEEE CEC is the IEEE Computational Intelligence Society (IEEE CIS)
flagship annual conference on evolutionary computation. The conference
promotes all aspects of theory, algorithm design and applications of
evolutionary computation and related fields focusing on mimicking nature
to solve problems. This year IEEE CEC has attracted 542 papers submitted
by 1460 authors from over 60 different countries. Each paper has a
minimum of 2 reviews, with the average of 3.4 reviews, done by 575
reviewers. Under the guidance of the Program Co-Chairs, Conflict of
Interest Chair and Technical Co-Chairs, and the help of special session
organizers, 324 papers of 910 authors were finally accepted, with an
acceptance rate of 6percent. All of them are included in the IEEE CEC 2021
program as oral presentations. The top ten countries where the
submitting authors come from are (in descending order): China, Brazil,
Japan, India, USA, Poland, Spain, New Zealand, Australia, and Canada.
The event will be attended by more than 470 participants. IEEE CEC 2021
highlights 24 tutorials given by experts in the most relevant and
emerging topics of evolutionary computation and related fields, 2
workshops, 2 panel discussions, 10 competitions, the CIS student branch
forum, and the mentoring program. The conference also attracted 63
special session proposals that led to 28 dedicated special sessions
organized by domain experts covering focused topics. For the cases where
scheduling constraints and the number of accepted papers did not allow
for a dedicated special session slot, the corresponding papers were
merged with an appropriate regular session with common theme. },
keywords = {},
doi = {doi:10.1109/CEC45853.2021.9504913},
  notes = {Also known as \cite{9504913}},
)


@INPROCEEDINGS(:2021:CEC,
%xplor 25 Aug 2021
  author = {},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Author Index},
year={2021},
  editor = {Yew-Soon Ong},
pages = {ii--xxi},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Presents an index of the authors whose articles are published
in the conference proceedings record.},
keywords = {},
doi = {doi:10.1109/CEC45853.2021.9504872},
  notes = {Also known as \cite{9504872}},
)


@INPROCEEDINGS(:2021:CEC,
%xplor 25 Aug 2021
  author = {},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Table of Contents},
year={2021},
  editor = {Yew-Soon Ong},
pages = {xxii--lvi},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Presents the table of contents/splash page of the proceedings
record.},
keywords = {},
doi = {doi:10.1109/CEC45853.2021.9504744},
  notes = {Also known as \cite{9504744}},
)


@INPROCEEDINGS(Zhang:2021:CEC,
%xplor 25 Aug 2021
  author = {Haotian Zhang and Jianyong Sun and Zongben Xu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Learning to Mutate for Differential Evolution},
year={2021},
  editor = {Yew-Soon Ong},
pages = {},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Adaptive parameter control and mutation operator selection are
two important research avenues in differential evolution (DE). Existing
works consider the two avenues independently. In this paper, we propose
to unify the two modules and develop a unified parameterized mutation
operator. With different settings of the parameters, different mutation
operators can be retrieved. Further, the settings of the parameters
closely relate to the control parameters of the DE. By determining the
parameters we can achieve adaptive parameter control and mutation
operator selection simultaneously. We propose to use a neural network to
output the parameters and learn the network parameter by the natural
evolution strategies algorithm under the consideration of modeling the
evolution process as a Markov Decision Process. Experimental results on
the CEC 2018 test suite show that the proposed method performs
significantly better than traditional DEs with different operators and
an advanced adaptive DE. We further analyze the time complexity and
population diversity of the proposed method. The analysis shows that our
method can achieve a balanced exploration and exploitation with a
properly learned network.},
keywords = {Training, Adaptation models, Sociology, Neural networks, Process control, Evolutionary computation, Markov processes, Differential Evolution, Adaptive Parameter Control, Adaptive Operator Selection, Markov Descision Process},
doi = {doi:10.1109/CEC45853.2021.9504990},
  notes = {Also known as \cite{9504990}},
)


@INPROCEEDINGS(Santucci:2021:CEC,
%xplor 25 Aug 2021
  author = {Valentino Santucci},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Is Algebraic Differential Evolution Really a Differential
Evolution Scheme?},
year={2021},
  editor = {Yew-Soon Ong},
pages = {9--16},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Algebraic Differential Evolution (ADE) is a recently
proposed combinatorial evolutionary scheme which mimics the behaviour of
the classical Differential Evolution (DE) in discrete search spaces
which can be represented as finitely generated groups. ADE has been
successfully applied to both permutation and binary optimization
problems. However, in the previous works, the relationship between ADE
and the classical continuous DE has been only intuitively sketched
without any theoretical or experimental proof. Here, we fill this gap by
providing both theoretical and experimental justifications proving that
ADE is a full-fledged generalization of DE which works across different
search spaces. First, we formally prove that there exists a concrete
implementation of ADE's algebraic operations converging to the classical
vector operations of DE, then we propose a real-vector implementation of
ADE and we experimentally prove that its behaviour is statistically
equivalent to DE. As conclusion, we also pave the way for further
applications of the original DE idea to mixed discrete/continuous search
spaces.},
keywords = {MIMICs, Evolutionary computation, Optimization, Differential Evolution, Algebraic Differential Evolution, Search spaces, Theoretical analysis},
doi = {doi:10.1109/CEC45853.2021.9504692},
  notes = {Also known as \cite{9504692}},
)


@INPROCEEDINGS(Takahama:2021:CEC,
%xplor 25 Aug 2021
  author = {Tetsuyuki Takahama and Setsuko Sakai},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Adaptive Differential Evolution Algorithm Utilizing Failure
Information and Success Information},
year={2021},
  editor = {Yew-Soon Ong},
pages = {17--24},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Differential Evolution (DE) has been successfully applied to a
variety of optimization problems. The performance of DE is affected by
two algorithm parameters of the scaling factor and the crossover rate.
Much research has been done in order to adaptively control the
parameters. One of the most successful studies on adaptive parameter
control is JADE, where the two parameter values are generated according
to two probability distributions which are tuned by the parameter values
in success cases. In this paper, we propose a new method that utilizes
not only success information but also failure information. A measure,
which indicates how much the pair of generated parameter values will
lead to failure, is defined. The pair is rejected probabilistically
according to the measure and a new pair is regenerated until a good pair
is generated. The effect of the proposed method is shown by solving
various benchmark problems.},
keywords = {Benchmark testing, Probabilistic logic, Probability distribution, Particle swarm optimization, Optimization, Genetic algorithms, differential evolution, failure information, adaptive parameter control},
doi = {doi:10.1109/CEC45853.2021.9504715},
  notes = {Also known as \cite{9504715}},
)


@INPROCEEDINGS(Ali:2021:CEC,
%xplor 25 Aug 2021
  author = {Ismail M. Ali and Hasan H. Turan and Ripon K. Chakrabortty and Sondoss Elsawah and Michael J. Ryan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Integrated Differential Evolution-based Heuristic Method for
Product Family Design Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {25--32},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Increases in demand for a greater variety of products help
companies gain more shares of growing competitive markets but, in
contrast, lead to an increase in production processes and, therefore,
higher costs and longer lead times. Although several techniques for
platform formations and assembly lines have been introduced to enable
more varieties of goods to be produced, this also makes a system more
complex and less cost-efficient. This paper proposes a differential
evolution (DE) approach that incorporates a new heuristic method,
improved solution representation and enhanced crossover and mutation
operators for solving the modular-based product family design problem in
a reconfigurable manufacturing system. The heuristic is applied to
repair some solutions in the initial population by replacing eligible
components with packages to provide near-optimal solutions in the
initial stage and enable DE to find the optimal solution quickly. The
proposed crossover is designed to further use the repaired solutions to
produce new individuals with better qualities. Finally, a case study of
a kettle family is conducted to validate this heuristic method, with the
experimental results showing that it saves 57.percent of the purchasing costs
of components and, on average, 41.3percent of setup costs compared with those
of median-joining phylogenetic network- and non-platform-based
heuristics. Moreover, the proposed DE achieves improved performances
with average errors of 63.3percent and 38.5percent from those of the standard
versions of DE and a genetic algorithm, respectively, in terms of the
total production costs of producing the same variants.},
keywords = {Sociology, Production, Evolutionary computation, Companies, Maintenance engineering, Phylogeny, Statistics, differential evolution, repairing method, reconfigurable manufacturing system, product family design},
doi = {doi:10.1109/CEC45853.2021.9504926},
  notes = {Also known as \cite{9504926}},
)


@INPROCEEDINGS(Wei:2021:CEC,
%xplor 25 Aug 2021
  author = {Jing Wei and Zuling Wang and Yangyan Xu and Ze Chen},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Differential Evolution with Multi-factor Ranking Based
Parameter Adaptation for Global Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {33--40},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The performance of differential evolution (DE) algorithm
depends critically on the setting of mutation factor F and crossover
rate CR. In this paper, a multi-factor ranking based parameter
adaptation scheme is proposed to properly set the value of F and CR. The
proposed adaptation scheme includes a parameter storage and distribution
mechanism. The parameter storage mechanism is used to produce a suitable
value by considering the multi-factor ranking scheme, which can help the
algorithm enhance its exploration ability in the early stage. The
parameter distribution mechanism is a layered parameter calculation
strategy, in which the individuals generate new parameters based on
multi-factor ranking properties. This mechanism is helpful to improve
the exploitation capability of DE. The performance of the proposed
method has been evaluated on CEC2014 and CEC2015 test suites and
compared with related methods. The results show our method could achieve
good performance and outperform related methods.},
keywords = {Sociology, Evolutionary computation, Aerospace electronics, Statistics, Optimization, Parameter control, differential evolution, optimization},
doi = {doi:10.1109/CEC45853.2021.9504838},
  notes = {Also known as \cite{9504838}},
)


@INPROCEEDINGS(Bai:2021:CEC,
%xplor 25 Aug 2021
  author = {Hao Bai and Changwu Huang and Xin Yao},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Adaptive Differential Evolution based on Exploration and
Exploitation Control},
year={2021},
  editor = {Yew-Soon Ong},
pages = {41--48},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Search operator design and parameter tuning are essential
parts of algorithm design. However, they often involve trial-and-error
and are very time-consuming. A new differential evolution (DE) algorithm
with adaptive exploration and exploitation control (AEEC-DE) is proposed
in this work to tackle this challenge. The proposed method improves the
performance of DE by automatically selecting trial vector generation
strategies (both mutation and crossover operators) and dynamically
generating the associated control parameter values. A probability-based
exploration and exploitation measurement is introduced to estimate
whether the state of each newly generated individual is in exploration
or exploitation. The state of historical individuals is used to assess
the exploration and exploitation capabilities of different generation
strategies and parameter values. Then, the strategies and parameters of
DE are adapted following the common belief that evolutionary algorithms
(EAs) should start with exploration and then gradually change into
exploitation. The performance of AEEC-DE is evaluated through
experimental studies on a set of test problems and compared with several
state-of-the-art adaptive DE variants.},
keywords = {Heuristic algorithms, Process control, Estimation, Evolutionary computation, Reinforcement learning, Benchmark testing, Kernel, Algorithm Configuration, Differential Evolution, Parameter Control, Exploration and Exploitation},
doi = {doi:10.1109/CEC45853.2021.9504876},
  notes = {Also known as \cite{9504876}},
)


@INPROCEEDINGS(Cordoba:2021:CEC,
%xplor 25 Aug 2021
  author = {Alejadro Tapia Cordoba and Alvaro Rodriguez {del Nozal} and Daniel Gutierrez Reina and Pablo {Millan Gata}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Genetic Algorithm To Optimize Penstocks For Micro-Hydro Power
Plants},
year={2021},
  editor = {Yew-Soon Ong},
pages = {49--56},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A Micro Hydropower Plant (MHPP) is a suitable and effective
mean to provide electric power to rural remote communities without
harming the environment. However, the lack of resources and technical
training in these communities frequently leads to designs based of rules
of thumb, compromising both the generation capacity and efficiency. This
work makes an attempt to address this problem developing a new tool to
design the layout of MHPP. The tool relies on a discrete topographic
survey of the terrain and makes use of a Genetic Algorithm (GA) to
optimize the installation layout, making it possible to explicitly
incorporate needs and constraints such as power supply requirement, cost
of the installation, available water flow, and layout feasibility in
accordance with the real terrain profile. The algorithm is applied to a
real scenario in a remote community in Honduras, demonstrating its
capability to optimize these kind of installations.},
keywords = {Training, Microhydro power, Power supplies, Layout, Decision making, Evolutionary computation, Tools, hydropower plant, renewable energy, optimization, genetic algorithm},
doi = {doi:10.1109/CEC45853.2021.9504994},
  notes = {Also known as \cite{9504994}},
)


@INPROCEEDINGS(Chitty:2021:CEC,
%xplor 25 Aug 2021
  author = {Darren M. Chitty},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Ant Colony Optimisation Inspired Crossover Operator for
Permutation Type Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {57--64},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Meta-heuristic methods are commonly applied to difficult
permutation type problems such as the Traveling Salesman Problem (TSP).
Genetic Algorithms (GA) and Ant Colony Optimisation (ACO) are two of the
most successful methods. However, a GA requires specialist crossover
operators for permutation problems to avoid repetition. This paper
presents a novel crossover operator, ACOX, inspired by ACO which can
essentially combine both meta-heuristic methods. When applied to a range
of TSP instances the ACOX crossover method demonstrates considerable
improvements over standard GA crossover operators. ACOX is able to
achieve solutions within percent of the optimal for TSP instances of several
thousand cities without using any local search methodologies.},
keywords = {Runtime, Search methods, Urban areas, Sociology, Traveling salesman problems, Genetics, Statistics, Ant Colony Optimisation, Genetic Algorithm, Traveling Salesman Problem},
doi = {doi:10.1109/CEC45853.2021.9504893},
  notes = {Also known as \cite{9504893}},
)


@INPROCEEDINGS(Mahmoudi-Nezhad:2021:CEC,
%xplor 25 Aug 2021
  author = {N. Hesam {Mahmoudi Nezhad} and M. Ghaffarian Niasar},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Algorithm Performance and the Influence of its Control
Parameters on the Optimization of Optical Lens Design},
year={2021},
  editor = {Yew-Soon Ong},
pages = {65--70},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={One of the major challenges in optical lens design is to
ascertain the lens system with the highest image quality. The image
quality of the lens system, which is a measure of the performance of the
lens, is a function of aberrations. This function is highly nonlinear
and leads to the presence of multiple local minima in the design
(optimization) landscape. Evolutionary algorithms, specifically Genetic
Algorithm, are receiving attention in this field as an efficient global
optimization techniques for multi variables and nonlinear objective
functions. However, to the best of our knowledge, studies are as yet
unavailable to provide an analysis on the performance of GA and the
influence of its tuning parameters on the optimization of these systems.
Our research has been conducted to supply such information and to
provide a guideline on using GA, in GA-aided optical lens designs. The
performance of GA has been investigated in a general group of three-lens
systems. It is shown that GA is an efficient optimization technique in
this field, while applying the suitable tuning parameters of GA is
crucial. It has been realized that Gaussian Mutation (Scale of 0.5),
combined with Heuristic Crossover, with a Crossover Fraction of 0.6, was
the option which yielded good (i.e. the challengeable practically
expected) results. However, any variation of these parameters may
prevent the system from ever reaching an optimal configuration.},
keywords = {Optical design, Linear programming, Optical imaging, Adaptive optics, Nonlinear optics, Optimization, Tuning, Optical lens design, Optimization, Genetic Algorithm (GA), GA Tuning Parameters},
doi = {doi:10.1109/CEC45853.2021.9504778},
  notes = {Also known as \cite{9504778}},
)


@INPROCEEDINGS(Luis:2021:CEC,
%xplor 25 Aug 2021
  author = {Samuel Yanes Luis and Federico {Peralta Samaniego} and Daniel Gutierrez Reina and Sergio {Toral Marin}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Sample-Efficiency Comparison Between Evolutionary Algorithms
and Deep Reinforcement Learning for Path Planning in an Environmental
Patrolling Mission},
year={2021},
  editor = {Yew-Soon Ong},
pages = {71--78},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={For water environmental monitoring tasks, the use of
Autonomous Surface Vehicles has been a very common option to substitute
human interaction and increase the efficiency and speed in the water
quality measuring process. This task requires an optimization of the
trajectories of the vehicle following a non-homogeneous interest
coverage criterion which is a hard optimization problem. This issue is
aggravated whenever the resolution of the water resource to be monitored
scales up. Since two of the preferred approaches for path planning of
autonomous vehicles in the literature are Evolutionary Algorithms and
Reinforcement Learning, in this paper, We compare the performance of
both techniques in a simulator of Ypacarai Lake in Asuncion (Paraguay).
The results show that the evolutionary approach is 5percent more efficient
for the lowest resolution but scales badly. Regarding the learning
stability and sparsity of the trajectory optimality, the Double Deep
Q-Learning algorithm has better convergence, but it appears to be less
robust than the evolutionary approach. Finally, in a generalization
analysis, the Deep Learning approach proves to be 3percent more effective in
reacting to scenario changes.},
keywords = {Stability criteria, Evolutionary computation, Reinforcement learning, Water quality, Robustness, Trajectory, Task analysis, autonomous vehicles, patrolling problem, path planning, deep reinforcement learning, evolutionary algorithms},
doi = {doi:10.1109/CEC45853.2021.9504864},
  notes = {Also known as \cite{9504864}},
)


@INPROCEEDINGS(de-Lima-Mendes:2021:CEC,
%xplor 25 Aug 2021
  author = {Raphael {de Lima Mendes} and Alexandre Henrick {da Silva Alves} and Matheus {de Souza Gomes} and Pedro {Luiz Lima Bertarini} and Laurence Rodrigues {do Amaral}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={ga{CNN:} Composing {CNNs} and {GAs} to Build an Optimized Hybrid
Classification Architecture},
year={2021},
  editor = {Yew-Soon Ong},
pages = {79--86},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Convolutional Neural Networks (CNN) are considered the gold
standard for Computer Vision Problems. However, finding the best
architecture for CNN often requires handcrafted design and domain
knowledge. On the other hand, Genetic Algorithms (GAs) have proven to be
an efficient technique to optimize a wide range of problems. Thus, in
this paper, we propose the gaCNN, a hybrid classification architecture
composed of a CNN and a GA. The gaCNN utilizes heterogeneous activation
functions to classify images, optimizing its hyperparameters and
activation functions automatically, regardless of the analyzed dataset.
The results show that gaCNN is able to identify good architectures. For
the Fashion MNIST dataset, the gaCNN obtained, as best accuracy, 93.7percent,
better than 9 of 13 compared classifiers. For the MNIST Handwritten
dataset, the gaCNN obtained, as best accuracy, 99.1percent, better than 12 of
16 classifiers.},
keywords = {Computer vision, Computer architecture, Evolutionary computation, Classification algorithms, Convolutional neural networks, Genetic algorithms, Artificial Intelligence, Machine Learning, Evolutionary Computation, Convolutional Neural Networks, Genetic Algorithms},
doi = {doi:10.1109/CEC45853.2021.9504850},
  notes = {Also known as \cite{9504850}},
)


@INPROCEEDINGS(Anh:2021:CEC,
%xplor 25 Aug 2021
  author = {Do Tuan Anh and Huynh {Thi Thanh Binh} and Nguyen Hoang Long and Ta {Bao Thang} and Su Simon},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Two-level Genetic Algorithm for Inter-domain Path Computation
under Node-defined Domain Uniqueness Constraints},
year={2021},
  editor = {Yew-Soon Ong},
pages = {87--94},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Recent years have witnessed an increment in the number of
network components communicating through many network scenarios such as
multi-layer and multi-domain, and it may result in a negative impact on
resource utilization. An urgent requirement arises for routing the
packets most efficiently and economically in large multi-domain
networks. In tackling this complicated area, we consider the
Inter-Domain Path Computation problem under Node-defined Domain
Uniqueness Constraint (IDPC-NDU), which intends to find the minimum
routing cost path between two nodes that traverses every domain at most
once. Owing to the NP-Hard property of the IDPC-NDU, applying
metaheuristic algorithms to solve this problem usually proves more
efficient. In like manner, this paper proposes a Two-level Genetic
Algorithm (PGA), where the first level determines the order of the
visited domains, and the second level finds the shortest path between
the two given nodes. Furthermore, to facilitate the finding process, a
method to minimize the search space and a new chromosome encoding that
would reduce the chromosome length to the number of domains are
integrated into this proposed algorithm. To evaluate the efficiency of
the proposal, experiments on various instances were conducted. The
results demonstrated that PGA outperforms other algorithms and gives
results no more than twice the optimal values.},
keywords = {Filtering, Optimized production technology, Evolutionary computation, Routing, Encoding, Resource management, Proposals, Multi-Domain, Path Computation Element, Evolutionary Algorithm},
doi = {doi:10.1109/CEC45853.2021.9504728},
  notes = {Also known as \cite{9504728}},
)


@INPROCEEDINGS(Gupta:2021:CEC,
%xplor 25 Aug 2021
  author = {Rashi Gupta and Satyasai Jagannath Nanda},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Solving Dynamic Many-objective {TSP} using {NSGA-III} equipped with
SVR-RBF Kernel Predictor},
year={2021},
  editor = {Yew-Soon Ong},
pages = {95--102},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Dynamic multi-objective TSP (DMTSP) finds extensive
applications in scheduling and routing problems. The task is challenging
due to the change in problem environment (arrangement and number of
cities) after certain time period. To solve this, in this manuscript a
new prediction based dynamic multi-objective optimization method termed
as Dynamic non-dominated sorting genetic algorithm III (DNSGA-III) is
proposed. This approach reuses the information obtained from previous
Pareto optimal sets (POS) to train prediction models. The prediction has
been carried out with SVR-RBF, SVR-Linear, polynomial interpolation and
cubic spline based prediction approaches and to determine new solutions
that are closer to the reference points. This significantly promote
population diversity, along with desired convergence. Performance of the
proposed DNSGA-III approach has been validated on four benchmark JY test
problems. Further a sixteen cities DMTSP problem with two objective
functions is solved using the proposed algorithm.},
keywords = {Heuristic algorithms, Urban areas, Predictive models, Benchmark testing, Prediction algorithms, Dynamic scheduling, Kernel, TSP, Dynamic Multi-objective Optimization Problems, non-dominated sorting, Pareto Optimal Set},
doi = {doi:10.1109/CEC45853.2021.9504966},
  notes = {Also known as \cite{9504966}},
)


@INPROCEEDINGS(Liu:2021:CEC,
%xplor 25 Aug 2021
  author = {Zhening Liu and Handing Wang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Improved Population Prediction Strategy for Dynamic
Multi-Objective Optimization Algorithms Using Transfer Learning},
year={2021},
  editor = {Yew-Soon Ong},
pages = {103--110},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Many real-world optimization problems have dynamic multiple
objectives and constrains, such problems are called dynamic
multi-objective optimization problems (DMOPs). Although many dynamic
multi-objective evolutionary algorithms (DMOEAs) have been proposed to
solve DMOPs, how to effectively track the optimal solutions in dynamic
environments is still a major challenge for dynamic multi-objective
optimization. Two classical DMOEAs, population prediction strategy (PPS)
and transfer learning based DMOEA (Tr-DMOEA) are validated to have great
performance because they integrate machine learning mechanism for
optimization. However, there are still some disadvantages in both
algorithms. In this paper, we propose a combined algorithm to make up
the respective disadvantages of PPS and Tr-DMOEA. Our algorithm retains
the prediction method of PPS considering sufficient historical
information. Then, we improve the prediction strategy in Tr-DMOEA to
further modify the solutions provided by PPS. These modified solutions
finally construct the initial population for optimization in the new
environment. The experiment results indicate that our algorithm has the
overall best performance comparing with PPS and Tr-DMOEA on the test
problems.},
keywords = {Machine learning algorithms, Heuristic algorithms, Sociology, Transfer learning, Evolutionary computation, Prediction methods, Prediction algorithms, Dynamic multi-objective optimization, evolutionary algorithm, transfer learning},
doi = {doi:10.1109/CEC45853.2021.9504877},
  notes = {Also known as \cite{9504877}},
)


@INPROCEEDINGS(Tan:2021:CEC,
%xplor 25 Aug 2021
  author = {Qingshan Tan and Changhe Li and Hai Xia and Sanyou Zeng and Shengxiang Yang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Novel Scalable Framework For Constructing Dynamic
Multi-objective Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {111--118},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Modeling dynamic multi-objective optimization problems (DMOPs)
has been one of the most challenging tasks in the field of dynamic
evolutionary optimization. Based on the analysis of the existing DMOPs,
several features widely existed in real-world applications are not taken
into account: different objectives may have different function models
and variables to be optimized; and the number of conflicting variables
should be independent from the number of objectives; the time-linkage
property is not considered. In order to overcome the above issues, a
novel framework for constructing DMOPs is proposed, where all objectives
can be designed independently, and the number of the conflicting
variables can be tuned by users. Moreover, it is easy to add new dynamic
features to this framework. Several classical dynamic multi-objective
optimization algorithms are tested on four scenarios, results show that
these characteristics are challenging for the existing algorithms.},
keywords = {Schedules, Analytical models, Heuristic algorithms, Object oriented modeling, Evolutionary computation, Benchmark testing, Dynamic scheduling, Dynamic multi-objective optimization, Bench-mark design, Time-linkage},
doi = {doi:10.1109/CEC45853.2021.9504961},
  notes = {Also known as \cite{9504961}},
)


@INPROCEEDINGS(Wu:2021:CEC,
%xplor 25 Aug 2021
  author = {Sheng-Hao Wu and Ke-Jing Du and Zhi-Hui Zhan and Hua Wang and Jun Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Historical Information-based Differential Evolution for Dynamic
Optimization Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {119--126},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Dynamic optimization problems (DOP) widely exist in many
application fields and remain a challenge. The multi-population
evolutionary computation approach is an efficient framework for solving
the DOPs. The key issue of the multi-population approach is how to
efficiently generate subpopulations in new environment by well reusing
historical information in past environments. However, most existing
works try to divide the population into small ones merely considering
the individuals' distribution, which may have the limitation that the
partition leads to the unbalanced computational resources allocation
among subpopulations. To address this limitation, we propose a
historical information-based differential evolution (HIDE) to
effectively solve the DOP. Firstly, a region-based subpopulation
initialization (RSI) strategy is proposed to generate multiple
subpopulations in the new environment in a balanced way. By initializing
multiple subpopulations in different regions of the search space, the
diversity of the population is enhanced, which is helpful to solve the
DOP with multiple peaks. Secondly, to fully make use of the found peaks
of the previous environments, an archive-based historical information
reuse (AHIR) strategy is put forward to manage and reuse the historical
information to guide the search in the new environment. Experiments have
been carried out on the moving peaks benchmark (MPB), which is a
commonly used DOP test suite. Experimental results show that the
proposed HIDE algorithm generally outperforms the classic DE algorithms
with different parameter settings and some existing competitive
state-of-the-art DOP approaches.},
keywords = {Heuristic algorithms, Sociology, Evolutionary computation, Benchmark testing, Search problems, Resource management, Statistics, Dynamic optimization problem, differential evolution, multi-population approach},
doi = {doi:10.1109/CEC45853.2021.9504965},
  notes = {Also known as \cite{9504965}},
)


@INPROCEEDINGS(Plump:2021:CEC,
%xplor 25 Aug 2021
  author = {Christina Plump and Bernhard J. Berger and Rolf Drechsler},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Improving Evolutionary Algorithms by Enhancing an Approximative
Fitness Function through Prediction Intervals},
year={2021},
  editor = {Yew-Soon Ong},
pages = {127--135},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Evolutionary algorithms are a successful application of
bio-inspired behaviour in the field of Artificial Intelligence.
Transferring mechanisms such as selection, mutation, and recombination,
evolutionary algorithms are capable of surmounting the disadvantages of
traditional methods. Adjusting an evolutionary algorithm to a specific
problem requires both, a good understanding of the problem and deep
knowledge of the effects of choosing one or another operator in the
algorithm. This becomes an especially difficult task when the fitness
function is not analytically given - that is, exists only as an
approximation, that is highly dependent on the present training data. We
propose using prediction intervals to modify the fitness function such,
that worse fitness values are less penalized if they occur in a poorly
fitted area. We evaluate this with an example from material sciences as
well as four standard benchmark algorithms for evolutionary algorithms
using a Support Vector Regression for training the approximative fitness
function and find that our approach outperforms the naive approximative
function.},
keywords = {Training, Support vector machines, Training data, Evolutionary computation, Benchmark testing, Prediction algorithms, Approximation algorithms},
doi = {doi:10.1109/CEC45853.2021.9504722},
  notes = {Also known as \cite{9504722}},
)


@INPROCEEDINGS(Bai:2021:CEC,
%xplor 25 Aug 2021
  author = {Wenlei Bai and Xinxin Zhu and Kwang Y. Lee},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Dynamic Optimal Power Flow Based on a Spatio-Temporal Wind Speed
Forecast Model},
year={2021},
  editor = {Yew-Soon Ong},
pages = {136--143},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={With large wind energy penetration to power grid, power system
operation has become more complex due to the intermittency of wind. For
an efficient operation of wind energy, accurate wind speed forecast is
in urgent need. Here, a statistical wind speed forecast model is
proposed which considers the spatial and temporal correlations in wind
speed and wind direction among geographically dispersed wind farms. Then
the forecast model was incorporated in the one-day ahead dynamic optimal
power flow for power system operation. Dynamic optimal power flow is a
highly non-linear and non-convex with large control variable
optimization problem. Modern heuristic optimization techniques have
proven their efficiency and robustness to such problem, so this work
focuses on a novel heuristic method, artificial bee colony. The original
artificial bee colony was modified in this work for dynamic
optimization. The forecast model has been verified by comparing with
actual wind speed. Several case studies are implemented on a modified
IEEE 30-bus system to verify the performances.},
keywords = {Wind energy, Wind speed, Power system dynamics, Evolutionary computation, Predictive models, Wind farms, Wind forecasting, Artificial bee colony (ABC), Data-driven forecast, Dynamic optimal power flow (DOPF), Evolutionary computation, Space-time model, Wind energy},
doi = {doi:10.1109/CEC45853.2021.9504847},
  notes = {Also known as \cite{9504847}},
)


@INPROCEEDINGS(Wu:2021:CEC,
%xplor 25 Aug 2021
  author = {Kai Eivind Wu and George Panoutsos},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A New Diversity Performance Indicator for Many-Objective
Optimisation Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {144--152},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Current performance indicators for assessing the diversity of
many-objective optimisation approximations are often underperforming as
the number of objectives increases, particularly for complex
optimisation problems. In this article, a new pure unary diversity
indicator is proposed, Inverse Ratio of Net Avertence angle (IRNA),
which is formulated by minimising the sum of the included angles between
approximation set and a set of reference vectors. It is achieved by
effectively rotating the reference vectors system in all dimensions
simultaneously with an optimised spatial angle. Any potential systematic
bias in included angles is removed, and the highest possible diversity
score of a solution set is obtained. Numerical results from evaluating
performance on synthetic solutions on a unit simplex plane and benchmark
functions of MaF show that the proposed performance indicator IRNA is
more sensitive to capturing diversity changes as the number of
objectives increases compared to other popular indicators.},
keywords = {Systematics, Sensitivity, Evolutionary computation, Benchmark testing, Approximation algorithms, Complexity theory, Optimization, Many-objective optimisation, Performance indicator, Diversity, Reference vectors, Benchmark testing},
doi = {doi:10.1109/CEC45853.2021.9504903},
  notes = {Also known as \cite{9504903}},
)


@INPROCEEDINGS(Zhou:2021:CEC,
%xplor 25 Aug 2021
  author = {Zhou Zhou and Zuling Wang and Tingting Pang and Jing Wei and Ze Chen},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Competition-Cooperation Evolutionary Algorithm with
Bidirectional Multi-population Local Search and Local Hypervolume-based
Strategy for Multi-objective Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {153--160},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper proposes a competition-cooperation algorithm with
bidirectional multi-population local search and local hypervolume-based
strategy (CCLS) to solve multi-objective optimization problems. In the
proposed method, a bidirectional multi-population local search is
devised and used to speed up convergence while keeping rich population
diversity. It searches along two completely opposite directions, and
adopts a unique strategy to perform replacement operation. Furthermore,
a local hypervolume-based strategy has been designed. It combines Pareto
dominance relation with hypervolume indicator to maintain population
diversity. The proposed algorithm is applied to the widely used
bi-objective and tri-objective test problems, and compared with related
methods. The results demonstrate that the proposed algorithm generally
outperforms related methods.},
keywords = {Sociology, Evolutionary computation, Search problems, Statistics, Optimization, Convergence, Multi-objective optimization, local search, hypervolume, competition-cooperation evolutionary algorithm},
doi = {doi:10.1109/CEC45853.2021.9504689},
  notes = {Also known as \cite{9504689}},
)


@INPROCEEDINGS(Nishihara:2021:CEC,
%xplor 25 Aug 2021
  author = {Kei Nishihara and Masaya Nakata},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Comparison of Adaptive Differential Evolution Algorithms on the
MOEA/D-DE Framework},
year={2021},
  editor = {Yew-Soon Ong},
pages = {161--168},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Existing works have reported that adaptive differential
evolution algorithms, i.e., adaptive DEs, improve the MOEA/D-DE
algorithm, but this result is limited to small-scale multi-objective
optimization problems. This paper compares four popular adaptive DEs on
the MOEA/D-DE framework to evaluate their scalability to the number of
decision variables and objectives. Specifically, we employ jDE, JADE,
EPSDE, and SaDE in this paper. Our experimental results provide the
following novel observations. MOEA/D-DE with JADE derives the best
average rank on small-scale problems. However, the performances of
MOEA/D-DE with JADE, EPSDE, and SaDE gradually degrade with the increase
of the problem scale. In contrast, jDE stably improves the performance
of MOEA/D-DE on large-scale problems employed in this paper (i.e., 11
objectives and 100 decision variables). Thus, we find a critical
tradeoff among adaptive DEs in terms of the scalability of the MOEA/D-DE
framework; a statistical adaption like JADE is suitable for small-scale
problems, but a randomization adaptation like jDE is effective with the
increase of the problem scale. Our results also suggest that
parameter-only adaptation can be suitable for MOEA/D-DE regardless of
the problem scale.},
keywords = {Scalability, Evolutionary computation, Optimization, algorithmic configuration adaptation, MOEA/D-DE, many-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504997},
  notes = {Also known as \cite{9504997}},
)


@INPROCEEDINGS(Liu:2021:CEC,
%xplor 25 Aug 2021
  author = {Tingrui Liu and Shenmin Song and Xin Li and Liguo Tan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Approximating Pareto Optimal Set by An Incremental Learning Model},
year={2021},
  editor = {Yew-Soon Ong},
pages = {169--176},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Combining a machine learning model within the search procedure
has shown great potentials in evolutionary multiobjective optimization
(EMO). The priori knowledge obtained from the property of Pareto optimal
set (PS) is a great help for reproducing high-quality offspring
solutions. However, the existing learning model in the framework of EMO
is also accompanied with a high computational cost resulted from its
iterative strategy or repetitive learning. To overcome this shortcoming,
the paper proposes to approximate the PS by an incremental learning
model. Specifically, it consists of two interdependent parts, i.e., a
learning module and a forgetting module. The basic idea is to take the
all new high-quality offspring solutions at the current evolution
iteration as a data stream, and incrementally train a model based on
Gaussian mixture models with the data stream to discover the manifold
structure of the PS and guide the evolutionary search. The learning
module is used to obtain the knowledge from the data stream in a batch
manner, while the forgetting module is applied to delete the information
from the relatively poor solution as is removed incrementally. The
proposed algorithm is employed to test suites, and the numerical
experiments demonstrates that the incremental learning model can help to
improve the algorithm performance with less computational cost compared
with the representative algorithms.},
keywords = {Manifolds, Computational modeling, Machine learning, Pareto optimization, Approximation algorithms, Data models, Computational efficiency},
doi = {doi:10.1109/CEC45853.2021.9504996},
  notes = {Also known as \cite{9504996}},
)


@INPROCEEDINGS(Xu:2021:CEC,
%xplor 25 Aug 2021
  author = {Zhouwu Xu and Jing Liu and Baihao Qiao and Yating Cao},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{MOEA/D} Using Dynamic Weight Vectors and Stable Matching Schemes
for the Deployment of Multiple Airships in the Earth Observing System},
year={2021},
  editor = {Yew-Soon Ong},
pages = {177--184},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The utilization of stratosphere in earth observing is getting
more attention. For maximizing the usage of stratosphere space, it is of
vital importance to deploy the platforms of which airship is the
representative, properly. Under such a background, a multiobjective
model for the deployment problem of multiple airships in the earth
observing system (MOM_DP_MAEOS), considering the number of covered tasks
and the total profits of the observable tasks, is designed. Furthermore,
a multiobjective evolutionary algorithm based on decomposition (MOEA/D)
with dynamic weight vectors (DW) and stable matching (STM) schemes
(MOEA/D_DW-STM) is proposed to optimize the deployment problem in this
paper. DW is applied in the process of updating the individuals in the
neighborhoods, playing the role of assisting local search. STM,
including a bi-direction selection process, is designed to overcome the
shortage that only the process of choosing solutions to subproblems
exists in MOEA/D. Besides, K-means clustering operator applied in the
initialization and the normalization operator for balancing the search
efforts of each objective are applied. A variety of experiments are
carried out on different kinds of benchmarks to verify the effectiveness
of MOEA/D_DW-STM and the rationality of MOM_DP_MAEOS. Compared with the
original MOEA/D and nondominated sorting genetic algorithm II (NSGA-II),
the designed algorithm obtains better Pareto fronts (PF) under the same
computational cost.},
keywords = {Earth, Heuristic algorithms, Atmospheric modeling, Earth Observing System, Terrestrial atmosphere, Evolutionary computation, Benchmark testing, Earth Observing System, Deployment Problem, Multiobjective Optimization Problem, Evolutionary Algorithm, Stable Matching},
doi = {doi:10.1109/CEC45853.2021.9504918},
  notes = {Also known as \cite{9504918}},
)


@INPROCEEDINGS(Lai:2021:CEC,
%xplor 25 Aug 2021
  author = {Guiyu Lai and Minhui Liao and Ke Li},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Empirical Studies on the Role of the Decision Maker in
Interactive Evolutionary Multi-Objective Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {185--192},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The interactive evolutionary multi-objective optimization
(IEMO) algorithms aim to learn and utilize the preference information
from the decision maker (DM) during the optimization process to guide
the search towards preferred solutions. In this paper, we are devoted to
figuring out the effects of interaction patterns, DM calls, preference
changes, and DM inconsistencies on the quality of the solutions
generated by the IEMO algorithms. The investigation is done in the
context of I-MOEA/D-PLVF algorithm, a recently proposed interactive
optimization algorithm based on MOEA/D.The experimental results indicate
that different interaction patterns and the number of DM calls do result
in significant impacts on the quality of the obtained solutions
generated by the IEMO algorithm used in our experiments. Meanwhile,
preference changes and DM inconsistencies in the process of interactions
will impose irreversibly negative effects on obtained solutions.},
keywords = {Time-frequency analysis, Computational modeling, Evolutionary computation, Search problems, Optimization, interactive multi-objective optimization, preference learning, decision maker},
doi = {doi:10.1109/CEC45853.2021.9504980},
  notes = {Also known as \cite{9504980}},
)


@INPROCEEDINGS(Marcelino:2021:CEC,
%xplor 25 Aug 2021
  author = {Carolina G. Marcelino and Lucas B. {de Oliveira} and Elizabeth F. Wanner and Carla A. D. M. Delgado and Silvia Jimenez-Fernandez and Sancho Salcedo-Sanz},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Hybrid Multiobjective Solution for the Short-term Hydro-power
Dispatch Problem: a Swarm Evolutionary Approach},
year={2021},
  editor = {Yew-Soon Ong},
pages = {193--200},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The unit dispatch problem is defined as the attribution of
operational values to each generation unit inside a hydro-power plant
(HPP), given some criteria such as the total power to be generated, or
the operational bounds of each unit. An optimal dispatch programming for
hydroelectric units in HPP provides a larger production of electricity,
with minimal water use. This paper presents an evolutionary approach to
optimize the multi-criteria electric dispatch problem in a general HPP,
based on a Multi-objective Evolutionary Swarm Hybridization (MESH)
algorithm. The proposed approach integrates mathematical models and
evolutionary swarm computation. The experimental analysis shows that the
proposed MESH algorithm is able to reach competitive results when
compared with classical evolutionary algorithms, the NGA-II and SPEA2
basing on ANOVA inference test. Results also show that the proposed MESH
is able to save a large amount of water in the energy production
process, supplying the requested load, and minimizing blackout risks and
generating a profit around $275,000 monthly.},
keywords = {Production, Evolutionary computation, Programming, Inference algorithms, Mathematical model, Analysis of variance},
doi = {doi:10.1109/CEC45853.2021.9504898},
  notes = {Also known as \cite{9504898}},
)


@INPROCEEDINGS(Biedrzycki:2021:CEC,
%xplor 25 Aug 2021
  author = {Rafal Biedrzycki and Kamil Kwiatkowski and Pawel Cichosz},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Compressor Schedule Optimization for a Refrigerated Warehouse
Using Metaheuristic Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {201--208},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper investigates the suitability of several
metaheuristic algorithms for the problem of compressor schedule
optimization for a refrigerated warehouse. A realistic simulator of such
a warehouse is used, based on domain knowledge, and tuned to match an
actual experimental cooling appliance. The problem consists in finding
an on-off sequence for the adopted optimization horizon and time step
that minimizes the energy cost while preserving cooling chamber
temperature constraints. To enable the application of metaheuristic
optimization algorithms, the problem has to be appropriately encoded.
Three different encoding schemes have been designed, suited to both
binary and continuous optimization methods. Several metaheuristic
algorithms known from the literature are used. Most of them deliver
solutions considerably better than a common-sense heuristic compressor
schedule. Interestingly, the classical genetic algorithm setup, as well
as a setup that was applied to a similar problem in prior research,
appear not to work well. The best results are achieved for an
alternative genetic algorithm configuration, determined by a series of
tuning experiments. Comparable results can be also obtained by the
IPOP-CMA-ES or PBIL algorithms, which do not require such extensive
tuning and may be preferred by practitioners.},
keywords = {Schedules, Cooling, Tariffs, Optimization methods, Turning, Linear programming, Noise measurement, refrigerated warehouse, problem encoding, binary search space},
doi = {doi:10.1109/CEC45853.2021.9504924},
  notes = {Also known as \cite{9504924}},
)


@INPROCEEDINGS(Campos:2021:CEC,
%xplor 25 Aug 2021
  author = {Italo Campos and Filipe Saraiva},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Reactive Multi-agent System for Self-healing in Smart
Distribution Grids},
year={2021},
  editor = {Yew-Soon Ong},
pages = {209--216},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In general, conventional power distribution systems lack
communication and automation capabilities, making almost impossible the
provision of features where self topology reconfiguration is necessary.
This scenario can be changed through the emergence of smart grids
technologies, where those capabilities are mandatory. An example of
feature possible by smart grids technology is the "self-healing", in
other words, the automatic loads restoration of the grid affected by a
fault. This paper describes and models a reactive multi-agent system
based in smart grid characteristics to provide the self-healing feature.
Following a sequence of steps, the modeled multi-agent system managed to
perform the network reconfiguration in a decentralized manner, using the
tie lines available in the grid. To validate the proposal, 20 different
computational simulations scenarios were implemented for two test
systems - 33 and 119 bus - defined in the literature. The obtained
results presented and discussed in this paper ensure that the approach
worked as expected.},
keywords = {Network topology, Computational modeling, Power distribution, Load shedding, Smart grids, Topology, Proposals, decentralized approach, multi-agent system, self-healing, simulation, smart grid},
doi = {doi:10.1109/CEC45853.2021.9505004},
  notes = {Also known as \cite{9505004}},
)


@INPROCEEDINGS(Huong:2021:CEC,
%xplor 25 Aug 2021
  author = {Tran Thi Huong and Le {Van Cuong} and Nguyen Ngoc Bao and Ngo {Minh Hai} and Huynh Thi {Thanh Binh}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Effective Partial Charging Scheme For Minimizing The Energy
Depletion And Charging Cost In Wireless Rechargeable Sensor Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {217--224},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Wireless Rechargeable Sensor Network has emerged as a
potential solution for the constrained energy problem in sensor networks
in recent years. The charging process has been employed to prolong the
sensor's lifetime. An effective charging algorithm requires
simultaneously optimize the charging path and charging time at each
charging location under the Mobile Charger's limited energy. The
existing methods, however, are generally lacking in the literature.
Moreover, most works are based on the assumption that Mobile Charger has
sufficient or infinite energy to visit and charge all sensors within
each charging cycle. This constraint leads to prolonging the waiting
charging time of energy-hurry sensors and unnecessary visiting of
energy-sufficient sensors. In this paper, we aim at minimizing energy
depletion and the charging cost of Mobile Charger in Wireless
Rechargeable Sensor Networks without the mentioned limitations above. We
first mathematically formulate the investigated problem as mixed integer
and linear programming. We propose a novel partial charging scheme based
on fuzzy logic and genetic algorithms to determine which sensors should
be charged in each cycle and optimize both charging paths and charging
time simultaneously. A range of experimental simulations is conducted to
demonstrate the effectiveness of our charging scheme. The simulation
results show our proposed algorithm's effectiveness compared to the
existing works concerning various performance metrics.},
keywords = {Wireless communication, Fuzzy logic, Measurement, Wireless sensor networks, Simulation, Evolutionary computation, Linear programming, Energy depletion, fuzzy logic, genetic algorithm, partial charging scheme, wireless rechargeable sensor networks},
doi = {doi:10.1109/CEC45853.2021.9504948},
  notes = {Also known as \cite{9504948}},
)


@INPROCEEDINGS(Almeida:2021:CEC,
%xplor 25 Aug 2021
  author = {Jose Almeida and Joao Soares and Bruno Canizes and Fernando Lezama and Mohammad Ali {Ghazvini Fotouhi} and Zita Vale},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolutionary Algorithms for Energy Scheduling under uncertainty
considering Multiple Aggregators},
year={2021},
  editor = {Yew-Soon Ong},
pages = {225--232},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The ever-increasing number of electric vehicles (EVs)
circulating on the roads and renewable energy production to achieve
carbon footprint reduction targets has brought many challenges to the
electrical grid. The increasing integration of distributed energy
resources (DER) in the grid is causing severe operational challenges,
such as congestion and overloading for the grid. Active management of
distribution network using the smart grid (SG) technologies and
artificial intelligence (AI) techniques can support the grid's operation
under such situations. Implementing evolutionary computational
algorithms has become possible using SG technologies. This paper
proposes an optimal day-ahead resource scheduling to minimize multiple
aggregators' operational costs in a SG, considering a high DER
penetration. The optimization is achieved considering three
metaheuristics (DE, HyDE-DF, CUMDANCauchy++). Results show that
CUMDANCauchy++ and HyDE-DF present the best overall results in
comparison to the standard DE.},
keywords = {Schedules, Uncertainty, Roads, Evolutionary computation, Scheduling, Smart grids, Mathematical model, aggregator, electric vehicles, energy resource management, evolutionary algorithms, smart grid, uncertainty},
doi = {doi:10.1109/CEC45853.2021.9504942},
  notes = {Also known as \cite{9504942}},
)


@INPROCEEDINGS(Lacerda:2021:CEC,
%xplor 25 Aug 2021
  author = {Eduardo Lacerda and Fernando Lezama and Joao Soares and Zita Vale},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{DSO} Contract Market for Demand Response Using Evolutionary
Computation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {233--240},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this article, a cost optimization problem in local energy
markets is analyzed considering fixed-term flexibility contracts between
the DSO and aggregators. The DSO procures flexibility while aggregators
of different types (e.g., conventional demand response or thermo-load
aggregators) offer the service. We solve the proposed model using
evolutionary algorithms based on the well-known differential evolution
(DE). First, a parameter-tuning analysis is done to assess the impact of
the DE parameters on the quality of solutions to the problem. Later,
after finding the best set of parameters for the "tuned" DE strategies,
we compare their performance with other self-adaptive parameter
algorithms, namely the HyDE, HyDE-DF, and vortex search algorithms.
Results show that with the identification of the best set of parameters
to be used for each strategy, the tuned DE versions lead to better
results than the other tested EAs. Overall, the algorithms are able to
find near-optimal solutions to the problem and can be considered an
alternative solver for more complex instances of the model.},
keywords = {Uncertainty, Computational modeling, Evolutionary computation, Load management, Regulation, Computational efficiency, Contracts, demand response, aggregator-DSO market, differential evolution, evolutionary computation},
doi = {doi:10.1109/CEC45853.2021.9504987},
  notes = {Also known as \cite{9504987}},
)


@INPROCEEDINGS(Luo:2021:CEC,
%xplor 25 Aug 2021
  author = {Xiaosong Luo and Zhao-Hui Sun and Siqi Qiu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Ant Colony System Based Drone Scheduling For Ship Emission
Monitoring},
year={2021},
  editor = {Yew-Soon Ong},
pages = {241--247},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Emission control area has been set up in many countries to
reduce the environmental impact of vessels' emissions. However, the
regulations for controlling emissions are frequently violated due to the
cost of high-quality fuel. Drones currently have become an accurate and
efficient way to monitor the vessels' emissions, which should be
properly scheduled to cover more and higher risk of violations when
facing a large number of vessels. In this paper, a scheduling model is
proposed to simulate the drone scheduling monitoring problem. Due to the
movement of vessels over time, the complexity of the model is too large
to be solved by classical optimization methods such as CPLEX. An ant
colony system algorithm is proposed to solve the scheduling problem of
drones. Our method is proved to be more effective and efficient when
facing a large number of vessels and drone stations in numerical
experiments.},
keywords = {Processor scheduling, Optimization methods, Inspection, Scheduling, Regulation, Numerical models, Fuels, Drone scheduling, Ship emission, Ant colony system, Optimization},
doi = {doi:10.1109/CEC45853.2021.9504944},
  notes = {Also known as \cite{9504944}},
)


@INPROCEEDINGS(Deng:2021:CEC,
%xplor 25 Aug 2021
  author = {Zhuang-Jie Deng and Liu-Yue Luo and Zhi-Hui Zhan and Jun Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Knowledge Embedding-Assisted Multi-Exemplar Learning Particle
Swarm Optimization for Traffic Signal Timing Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {248--255},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Traffic signal timing optimization (TSTO) has aroused
extensive attention, which aims to optimize the signal timing to improve
the service capability of intersections. In recent years, particle swarm
optimization (PSO) has played a vital role in optimization field.
However, the PSO may get stuck in the local optimum when handling TSTO.
In this paper, we propose a multi-exemplar learning PSO (MEL-PSO)
algorithm, which enhances the exploration capability of the algorithm by
letting particles have more opportunities to learn from more exemplars.
Moreover, a knowledge embedded solution generating (KESG) strategy is
proposed by exploiting the characteristic of input traffic volume
distribution as pre-knowledge, which helps MEL-PSO generate an initial
population covering promising search space. Furthermore, in order to
make MEL-PSO suitable for different kinds of saturation situations of
the intersection, we adopt multiple indicators to measure the
performance of the signal timing scheme. Comparison experiments for
validating the performance of MEL-PSO are carried out on a single
intersection in both undersaturated and oversaturated traffic flow
conditions. Experimental results show that MEL-PSO outperforms the
classic numerical timing method, random timing method, and some other
PSO-based algorithms.},
keywords = {Sociology, Timing, Statistics, Particle swarm optimization, Optimization, Traffic signal timing optimization, particle swarm optimization, multi-exemplar learning, knowledge embedding},
doi = {doi:10.1109/CEC45853.2021.9504703},
  notes = {Also known as \cite{9504703}},
)


@INPROCEEDINGS(Takahashi:2021:CEC,
%xplor 25 Aug 2021
  author = {Kenjiro Takahashi and Yoshikazu Fukuyama and Shuhei Kawaguchi and Takaomi Sato},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Optimal Production Scheduling using a Production Simulator by
Modified Brain Storm Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {256--263},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper proposes optimal production scheduling using a
practical production simulator and modified brain storm optimization
(MBSO). So far, in a production scheduling research area, for the
purpose of practical applications, various practical performance
measures and constraints have been proposed. Recently, production
scheduling researches in consideration of energy consumptions and CO2
emissions have been conducted. However, in the various production
scheduling researches, each research only treats practical performance
measures, practical constraints, energy consumptions, and CO2 emissions
partially. Hence, no researches have considered the various practical
performance measures and constraints, energy consumptions, and CO2
emissions totally. In this paper, a production schedule optimization
framework is proposed using a production simulator considering
constraints that cannot be mathematically formulated. It is verified
that the proposed MBSO based method can find higher quality solutions
than the conventional PSO and DEEPSO based methods. It is also verified
that there is a significant difference among the conventional PSO and
DEEPSO based methods, and the proposed MBSO based method with 0.05
significant level using the Friedman test.},
keywords = {Energy consumption, Schedules, Storms, Processor scheduling, Area measurement, Energy measurement, Production, optimal production scheduling, production simulator, modified brain storm optimization, CO2 emissions},
doi = {doi:10.1109/CEC45853.2021.9504730},
  notes = {Also known as \cite{9504730}},
)


@INPROCEEDINGS(Smirnov:2021:CEC,
%xplor 25 Aug 2021
  author = {Egor Smirnov and Sergei Kudinov},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Using a Genetic Algorithm for Planning Interesting Tourist Routes
in the City on the Basis of Open Street Map Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {264--271},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Walking is a good way to get to know a new district or city.
However, it may be difficult for a tourist not familiar with the area to
choose a route for a walk because they do not know, what places of
interest there are nearby. This is especially true for areas unpopular
with tourists and not mentioned in guidebooks. But even if attractions
are listed there, it can be challenging to think of the best route to
see them all. In this case, a special navigation service can help, which
is able to build more interesting walking routes rather than shorter
ones (as existing popular navigation services do). Such routes may be
longer but they include local places of interest, green areas and
embankments. The article proposes an algorithm that uses data on places
of interest from Open Street Map and applies a modernized A* algorithm
as well as a genetic algorithm to build such routes. Creation of two
route options is considered: a route between two different points (if
the tourist wants to reach a specific destination while seeing some
places of interest along the way) and a circular route returning to the
starting point in a specified time (e.g., if the tourist wants to walk
around the hotel or a train station).},
keywords = {Legged locomotion, Navigation, Urban areas, Evolutionary computation, Planning, Genetic algorithms, Tourism, navigation, path finding, places of interest, walking routes, genetic algorithms, Open Street Map},
doi = {doi:10.1109/CEC45853.2021.9504741},
  notes = {Also known as \cite{9504741}},
)


@INPROCEEDINGS(Robinson:2021:CEC,
%xplor 25 Aug 2021
  author = {Demelza Robinson and Qi Chen and Bing Xue and Daniel Killeen and Sara Fraser-Miller and Keith C Gordon and Indrawati Oey and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Algorithm for Feature and Latent Variable Selection for
Nutrient Assessment in Horticultural Products},
year={2021},
  editor = {Yew-Soon Ong},
pages = {272--279},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Vibrational spectroscopy can be used for rapid determination
of chemical quality markers in horticultural produce to improve quality
control, optimize harvest times and maximize profits. Most commonly,
spectral data are calibrated against chemical reference data (acquired
using traditional, slower analytical methods) using partial least
squares regression (PLSR). However, predictive performance of PLSR can
be limited by the small number of instances, high dimensionality and
collinearity of spectroscopic data. Here, a new genetic algorithm (GA)
for PLSR feature and latent variable selection is proposed to predict
concentrations of 18 important bioactive components across three New
Zealand horticultural products from infrared, near-infrared and Raman
spectral data sets. Models generated using the GA-enhanced PLSR method
have notably better generalization and are less complex than the
standard PLSR method. GA-enhanced PLSR models are produced from each
spectroscopic data set individually, and from a data set that combines
all three techniques.},
keywords = {Spectroscopy, Input variables, Biological system modeling, Computational modeling, Quality control, Evolutionary computation, Data models, Genetic Algorithm, Feature Selection, Partial Least Squares Regression, Vibrational Spectroscopy},
doi = {doi:10.1109/CEC45853.2021.9504794},
  notes = {Also known as \cite{9504794}},
)


@INPROCEEDINGS(Hirsch:2021:CEC,
%xplor 25 Aug 2021
  author = {Laurence Hirsch and Alessandro Di Nuovo and Prasanna Haddela},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Document Clustering with Evolved Single Word Search Queries},
year={2021},
  editor = {Yew-Soon Ong},
pages = {280--287},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We present a novel, hybrid approach for clustering text
databases. We use a genetic algorithm to generate and evolve a set of
single word search queries in Apache Lucene format. Clusters are formed
as the set of documents matching a search query. The queries are
optimized to maximize the number of documents returned and to minimize
the overlap between clusters (documents returned by more than one query
in a set). Optionally, the number of clusters can be specified in
advance, which will normally result in an improvement in performance.
Not all documents in a collection are returned by any of the search
queries in a set, so once the search query evolution is completed a
second stage is performed whereby a KNN algorithm is applied to assign
all unassigned documents to their nearest cluster. We describe the
method and compare effectiveness with other well-known existing systems
on 8 different text datasets. We note that search query format has the
qualitative benefits of being interpretable and providing an explanation
of cluster construction.},
keywords = {Databases, Clustering algorithms, Evolutionary computation, Media, Genetic algorithms, document clustering, search query, genetic algorithm, Lucene},
doi = {doi:10.1109/CEC45853.2021.9504770},
  notes = {Also known as \cite{9504770}},
)


@INPROCEEDINGS(Mandal:2021:CEC,
%xplor 25 Aug 2021
  author = {Ranju Mandal and Basim Azam and Brijesh Verma and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Deep Learning Model with {GA-based} Visual Feature Selection and
Context Integration},
year={2021},
  editor = {Yew-Soon Ong},
pages = {288--295},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Deep learning models have been very successful in computer
vision and image processing applications. Since its inception,
Convolutional Neural Network (CNN)-based deep learning models have
consistently outperformed other machine learning methods on many
significant image processing benchmarks. Many top-performing methods for
image segmentation are also based on deep CNN models. However, deep CNN
models fail to integrate global and local context alongside visual
features despite having complex multi-layer architectures. We propose a
novel three-layered deep learning model that learns independently global
and local contextual information alongside visual features, and visual
feature selection based on a genetic algorithm. The novelty of the
proposed model is that One-vs-All binary class-based learners are
introduced to learn Genetic Algorithm (GA) optimized features in the
visual layer, followed by the contextual layer that learns global and
local contexts of an image, and finally the third layer integrates all
the information optimally to obtain the final class label. Stanford
Background and CamVid benchmark image parsing datasets were used for our
model evaluation, and our model shows promising results. The empirical
analysis reveals that optimized visual features with global and local
contextual information play a significant role to improve accuracy and
produce stable predictions comparable to state-of-the-art deep CNN
models.},
keywords = {Deep learning, Visualization, Image segmentation, Computational modeling, Benchmark testing, Feature extraction, Convolutional neural networks, Image Parsing, Deep Learning, Genetic Algorithm, Scene Understanding, Semantic Segmentation},
doi = {doi:10.1109/CEC45853.2021.9504753},
  notes = {Also known as \cite{9504753}},
)


@INPROCEEDINGS(Nguyen:2021:CEC,
%xplor 25 Aug 2021
  author = {Trung B. Nguyen and Will N. Browne and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Constructing Complexity-efficient Features in {XCS} with Tree-based
Rule Conditions},
year={2021},
  editor = {Yew-Soon Ong},
pages = {296--303},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A major goal of machine learning is to create techniques that
abstract away irrelevant information. The generalisation property of
standard Learning Classifier Systems (LCSs) removes such information at
the feature level but not at the feature interaction level. Code
Fragments (CFs), a form of tree-based programs, introduced feature
manipulation to discover important interactions, but they often contain
irrelevant information, which causes structural inefficiency. XOF is a
recently introduced LCS that uses CFs to encode building blocks of
knowledge about feature interaction. This paper aims to optimise the
structural efficiency of CFs in XOF. We propose two measures to improve
constructing CFs to achieve this goal. Firstly, a new CF-fitness update
estimates the applicability of CFs to the problem while also considering
the structural complexity. The second measure is a niche-based method
for generating CFs. These approaches were tested on Even-parity and
Hierarchical problems, which require highly complex combinations of
input features to capture the data patterns. The results show that the
proposed methods significantly increase the structural efficiency of
CFs, which is estimated by the rule "generality rate". This results in
faster learning performance in the Hierarchical Majority-on problem.
Furthermore, a user-set depth limit for CF generation is not needed as
the learning agent will not adopt higher-level CFs once optimal CFs are
constructed.},
keywords = {Machine learning, Knowledge representation, Evolutionary computation, Complexity theory, Reliability, Task analysis, Standards, LCS, XCS, Code Fragments, XOF},
doi = {doi:10.1109/CEC45853.2021.9504784},
  notes = {Also known as \cite{9504784}},
)


@INPROCEEDINGS(Tadokoro:2021:CEC,
%xplor 25 Aug 2021
  author = {Masakazu Tadokoro and Hiroyuki Sato and Keiki Takadama},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{XCS} with Weight-based Matching in {VAE} Latent Space and Additional
Learning of High-Dimensional Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {304--310},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, we propose MVN-ELSDeCS, which is a combination
of VAE, a dimensionality reduction network, and MVN-XCSR, an XCSR
extended to a distributional representation. In addition, additional
learning of high-dimensional data with XCS is performed to reduce the
information loss in learning caused by dimensionality reduction. We
applied the proposed method to the benchmark problem of 10-class
classification of handwritten digit images, and the experimental results
have the following implications: 1) MVN-XCSR, which is a component of
MVN-ELSDeCS, not only shows higher classification performance from the
early stage of training in the dimensionally compressed latent space,
but also 2) the reconstructed rules generated by MVN-ELSDeCS shows
higher classification performance for the original high-dimensional
data. Furthermore, 3) by applying additional learning with XCS to the
reconstructed rules, the classification accuracy of rules for the
10-class classification task was significantly improved.},
keywords = {Dimensionality reduction, Training, Image coding, Evolutionary computation, Benchmark testing, Task analysis, Image reconstruction, learning classifier system, VAE, XCS, Weight-based matching, MNIST problem},
doi = {doi:10.1109/CEC45853.2021.9504909},
  notes = {Also known as \cite{9504909}},
)


@INPROCEEDINGS(Shiraishi:2021:CEC,
%xplor 25 Aug 2021
  author = {Hiroki Shiraishi and Masakazu Tadokoro and Yohei Hayamizu and Yukiko Fukumoto and Hiroyuki Sato and Keiki Takadama},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Increasing Accuracy and Interpretability of High-Dimensional
Rules for Learning Classifier System},
year={2021},
  editor = {Yew-Soon Ong},
pages = {311--318},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
  abstract = {This paper proposes ELPSDeCS (Encoding, Learning, 'Plausibl},
Sampling, and Decoding Classifier System) by extending ELSDeCS
(Encoding, Learning, Sampling, and Decoding Classifier System) to
increase both the accuracy and interpretability of the generated
classifiers which matches the high-dimensional input such as images. The
experimental results on the complex multi-class classification problem
of the handwritten numerals show that both the accuracy and
interpretability of ELPSDeCS are higher than that of ELSDeCS.},
keywords = {Image coding, Impedance matching, Reinforcement learning, Evolutionary computation, Decoding, Data mining, Task analysis, data mining, XCS, Auto-Encoders, interpretability},
doi = {doi:10.1109/CEC45853.2021.9504733},
  notes = {Also known as \cite{9504733}},
)


@INPROCEEDINGS(Leite:2021:CEC,
%xplor 25 Aug 2021
  author = {Gabriel M. C. Leite and Carolina G. Marcelino and Elizabeth F. Wanner and Carlos E. Pedreira and Silvia Jimenez-Fernandez and Sancho Salcedo-Sanz},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Pattern Classification Applying Neighbourhood Component Analysis
and Swarm Evolutionary Algorithms: A Coupled Methodology},
year={2021},
  editor = {Yew-Soon Ong},
pages = {319--326},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this work we present a pattern classification approach
coupling the Neighbourhood Component Analysis (NCA) classifier with the
Canonical Differential Evolutionary Particle Swarm Optimization
(C-DEEPSO). The standard NCA uses the conjugate gradient method to
minimize the classification error. Here we propose an approach using the
C-DEEPSO instead. In the experimental design, the coupled approach is
applied to 20 benchmark data sets, and its performance is compared with
the standard NCA using the conjugate gradient. The experimental analysis
shows the usage of an evolutionary approach to enhance the performance
of a machine learning algorithm can be competitive when compared to
well-known iterative optimization techniques, and even outperform them
in some problems. A real-world problem classifying cyber-attacks to an
industrial control system of gas pipelines is also solved by the
proposed approach. The results obtained indicate the proposed approach
can successfully identify possible cyber-attacks to the control system.
In this way, the NCA coupled to C-DEEPSO can work as an Intrusion
Detection Systems (IDS), being able to guarantee an acceptable security
level.},
keywords = {Machine learning algorithms, Industrial control, Pipelines, Pattern classification, Intrusion detection, Iterative algorithms, Security, Machine Learning, Classification Problems, Neighbourhood Component Analysis, Evolutionary Algorithms, Intrusion Detection Systems},
doi = {doi:10.1109/CEC45853.2021.9504702},
  notes = {Also known as \cite{9504702}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {Junyuan Chen and Jing Liu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Memetic Algorithm for Optimizing Inter-links to Enhance the
Robustness of Interdependent Networks Against Malicious Attacks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {327--334},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In real life, independent networks exist widely and their
robustness has received significant attention. When an interdependent
network faces attacks, the failure can propagate along inter-links to
the whole network and the damage caused by malicious attacks is more
serious. However, in terms of designing the robust inter-links of
interdependent networks, most of the existing strategies target at
random attacks and make little use of the inner structure of
interdependent networks. In this paper, we propose a specific memetic
algorithm (MA) for optimizing the inter-links to enhance the robustness
of interdependent networks against malicious attacks, termed as
MA-RINint. Moreover, we design an effective crossover operator based on
retaining the inter-links of adjacent nodes and a local search operator
according to the degree of nodes. In the experiments, we compare the
proposed algorithm with five existing methods on three types of
interdependent networks with different sizes, and the experimental
results show that MA-RINint performs excellently in enhancing the
robustness of interdependent networks.},
keywords = {Memetics, Computational modeling, Evolutionary computation, Genetics, Robustness, Optimization, Faces, interdependent networks, robustness, memetic algorithm, malicious attacks},
doi = {doi:10.1109/CEC45853.2021.9504787},
  notes = {Also known as \cite{9504787}},
)


@INPROCEEDINGS(Neri:2021:CEC,
%xplor 25 Aug 2021
  author = {Ferrante Neri and Yuyang Zhou},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Covariance Pattern Search with Eigenvalue-determined Radii},
year={2021},
  editor = {Yew-Soon Ong},
pages = {335--342},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Effective implementations of Memetic Algorithms often
integrate, within their design, problem-based pieces of information.
When no information is known, an efficient MA can still be designed
after a preliminary analysis of the problem. This approach is usually
referred to as Fitness Landscape Analysis (FLA). This paper proposes a
FLA technique to analyse the epistasis of continuous optimisation
problems and estimate those directions, within a multi-dimensional
space, associated with maximum and minimum directional derivatives. This
estimation is achieved by making use of the covariance matrix associated
with a distribution of points whose objective function value is below
(in case of minimisation) a threshold. The eigenvectors and eigenvalues
of the covariance matrix provide important pieces of information about
the geometry of the problem and are then used to design a memetic
operator that is a local search belonging to the family of generalised
Pattern Search. A restarting mechanism enables a progressive
characterisation of the fitness landscape. Numerical results show that
the proposed approach successfully explore ill-conditioned basins of
attractions and outperforms the standard pattern search as well as a
pattern search recently proposed in the literature and partially based
on a similar design logic. The proposed local search based on FLA also
displays a performance competitive with that of other types of local
search.},
keywords = {Memetics, Geometry, Search problems, Minimization, Linear programming, Eigenvalues and eigenfunctions, Space exploration, Memetic Algorithms, Fitness Landscape Analysis, Pattern Search, Covariance Matrix, Local Search},
doi = {doi:10.1109/CEC45853.2021.9505002},
  notes = {Also known as \cite{9505002}},
)


@INPROCEEDINGS(Ali:2021:CEC,
%xplor 25 Aug 2021
  author = {Hanan Ba Ali and Adam Roman},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Designing Urban Transit Network using Memetic Algorithm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {343--350},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Urban transit network design (UTND) problem represents a
challenge in designing routes, with a trade-off between serving
passengers and operators benefits. In this study, a Memetic Algorithm
(MA) is proposed to solve the UTND problem. The algorithm uses the hill
climbing local search (HCLS) algorithm as an additional operator for
Genetic Algorithm (GA) to improve routes construction during the global
search. The proposed method consists of two phases. In the first phase,
a set of solutions (transit network designs) is generated as an initial
population for MA, where each solution consists of a set of routes. The
predefined set of solutions satisfies the constraints such as route
length or number of routes, and requirements like lack of loops, and
that all nodes are covered by at least one route. In the second phase,
the suggested Memetic Algorithm (MA) is used to generate all possible
solutions from the predefined set. The MA tries to find the best
structured solution that represents the flawless transit network. The
proposed MA is applied on the widely examined benchmark problems: Mandl
and Mumford networks. The experiment results show that the suggested MA
provides significant improvements in terms of the direct trip percentage
and average travel time compared to the previous studies.},
keywords = {Memetics, Sociology, Evolutionary computation, Benchmark testing, Statistics, Genetic algorithms, Transit Network Design, Local Search Algorithms, Evolutionary Algorithms},
doi = {doi:10.1109/CEC45853.2021.9504756},
  notes = {Also known as \cite{9504756}},
)


@INPROCEEDINGS(Osaba:2021:CEC,
%xplor 25 Aug 2021
  author = {Eneko Osaba and Esther Villar-Rodriguez and Izaskun Oregi and Aitor Moreno-Fernandez-de-Leceta},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Hybrid Quantum Computing - Tabu Search Algorithm for Partitioning
Problems: Preliminary Study on the Traveling Salesman Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {351--358},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Quantum Computing is considered as the next frontier in
computing, and it is attracting a lot of attention from the current
scientific community. This kind of computation provides to researchers
with a revolutionary paradigm for addressing complex optimization
problems, offering a significant speed advantage and an efficient search
ability. Anyway, Quantum Computing is still in an incipient stage of
development. For this reason, present architectures show certain
limitations, which have motivated the carrying out of this paper. In
this paper, we introduce a novel solving scheme coined as hybrid Quantum
Computing - Tabu Search Algorithm. Main pillars of operation of the
proposed method are a greater control over the access to quantum
resources, and a considerable reduction of non-profitable accesses. To
assess the quality of our method, we have used 7 different Traveling
Salesman Problem instances as benchmarking set. The obtained outcomes
support the preliminary conclusion that our algorithm is an approach
which offers promising results for solving partitioning problems while
it drastically reduces the access to quantum computing resources. We
also contribute to the field of Transfer Optimization by developing an
evolutionary multiform multitasking algorithm as initialization method.},
keywords = {Quantum computing, Evolutionary computation, Computer architecture, Traveling salesman problems, Benchmark testing, Search problems, Multitasking, Quantum Computing, Metaheuristic Optimization, Traveling Salesman Problem, Transfer Optimization, DWAVE},
doi = {doi:10.1109/CEC45853.2021.9504923},
  notes = {Also known as \cite{9504923}},
)


@INPROCEEDINGS(Bidgoli:2021:CEC,
%xplor 25 Aug 2021
  author = {Azam Asilian Bidgoli and Shahryar Rahnamayan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Memetic Differential Evolution Using Coordinate Descent},
year={2021},
  editor = {Yew-Soon Ong},
pages = {359--366},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Differential Evolution (DE) is one of the well-established
population-based optimization algorithms which has received a lot of
attention regarding its potential to solve complex optimization
problems. However, DE is capable to explore a huge search space in its
early run phase, called exploration phase, its weakness in exploitation
avoids local refinement of the promising shrunk region. Therefore,
employing a local search can be an efficient strategy to improve the
search performance of DE via accelerating of fine tuning phase. This
paper purposes an effective Memetic DE algorithm using a well-known
single-solution-based optimization method, i.e., Coordinate Descent (CD)
algorithm. Local coordinate search is applied on the promising region
resulted by top ranked individuals selected from the final population of
DE. The proposed method updates the value of each coordinate iteratively
by evaluating the sampled points from the local region to improve the
resulted candidate solution. Since coordinate search algorithm shrinks
the region rapidly, it requires a very small portion of the
computational budget to find the optimal coordinates' value. In order to
evaluate the proposed Memetic DE, several experiment series are
conducted on functions of CEC-2017 benchmark for different number of
dimensions (i.e., D=30, 50, and 100). Results clearly indicate that the
utilized local coordinate search improves the quality of resulted
solution by DE significantly using a very low computational budget,
i.e., 20xD.},
keywords = {Memetics, Sociology, Optimization methods, Evolutionary computation, Benchmark testing, Statistics, Tuning, Differential Evolution (DE), Coordinate Descent, Local Search, Population-based Algorithms, Metaheuristics, Memetic Algorithm},
doi = {doi:10.1109/CEC45853.2021.9504762},
  notes = {Also known as \cite{9504762}},
)


@INPROCEEDINGS(Pan:2021:CEC,
%xplor 25 Aug 2021
  author = {Zi-Xiao Pan and Ling Wang and Jing-Fang Chen and Yu-Ting Wu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Novel Evolutionary Algorithm with Adaptation Mechanism for
Fuzzy Permutation Flow-Shop Scheduling},
year={2021},
  editor = {Yew-Soon Ong},
pages = {367--374},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={As a scheduling problem with wide application backgrounds, the
flow-shop scheduling problem (FSP) in deterministic cases has attracted
much attention. However, the neglecting of uncertainty will greatly
diminish the application value of scheduling results, which makes it
necessary to incorporate the uncertainty in the FSP. In this paper, a
fuzzy permutation flow-shop scheduling problem (FPFSP) is considered and
a novel evolutionary algorithm with adaptation mechanism (AMEA) is
proposed to minimize fuzzy makespan. In the initialization phase, two
initialization strategies based on the NEH heuristic are proposed to
improve the quality of initial population. In the evolution phase, to
enhance the exploitation, multiple local search operators are conducted
in a collaborative way where the utilizations of operators are adjusted
adaptively according to the feedback of their performances; besides, to
save computing resources and balance the exploration and exploitation,
the population size is adjusted adaptively with the number of
generations. Benchmark instances are generated to evaluate the
performance of the AMEA. The experimental results and statistical
comparisons show that the proposed algorithm has great advantages in
solving the FPFSP.},
keywords = {Job shop scheduling, Uncertainty, Processor scheduling, Sociology, Collaboration, Evolutionary computation, Reinforcement learning, fuzzy scheduling, flow-shop scheduling, adaptation mechanism, evolutionary algorithm},
doi = {doi:10.1109/CEC45853.2021.9504870},
  notes = {Also known as \cite{9504870}},
)


@INPROCEEDINGS(Nogueira-Lorena:2021:CEC,
%xplor 25 Aug 2021
  author = {Luiz Henrique {Nogueira Lorena} and Luiz Antonio {Nogueira Lorena} and Antonio Augusto Chaves},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Biased Random-key Genetic Algorithm with a Local Search
Component for the Optimal Bucket Order Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {375--382},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Aggregating ranks into a consensus is an important task
applied in different fields of science. This paper deals with a specific
variation that aggregate ranks into a consensus considering ties between
its elements. This approach is more flexible and meaningful for modeling
some circumstances where a strict order is considered too restrictive. A
ranking considering ties is also known as a bucket order in literature,
and the problem that considers the rank aggregation of bucket orders is
defined as the Optimal Bucket Order Problem (OBOP). It is an NP-hard
problem, hence several heuristics have been proposed in the literature.
The current state-of-the-art results for this problem were achieved
through an Evolution Strategy (ES) metaheuristic. This paper proposes
the application of the adaptive Biased Random-key Genetic Algorithm
(A-BRKGA) with Variable Neighborhood Descent (VND) as a local search to
solve it. The A-BRKGA is a metaheuristic with on-line parameter control,
in which the strategy for parameter tuning is based on deterministic
rules and self-adaptive schemes. The proposed approach was compared with
ES in 152 instances, improving the fitness of the best solutions in
35.5percent of the instances, providing better average solutions for 70.3percent,
and equal results for the remaining instances.},
keywords = {NP-hard problem, Aggregates, Evolutionary computation, Search problems, Task analysis, Tuning, Genetic algorithms, Genetic algorithm, Optimal bucket order problem, Rank aggregation, Consensus ranking},
doi = {doi:10.1109/CEC45853.2021.9504985},
  notes = {Also known as \cite{9504985}},
)


@INPROCEEDINGS(Shimazu:2021:CEC,
%xplor 25 Aug 2021
  author = {Sho Shimazu and Isao Ono},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Evolutionary Algorithm Taking Account of Epistasis among
Parameters for Black-Box Discrete Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {383--390},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We propose an evolutionary algorithm that takes account of
epistasis among parameters for black-box discrete optimization problems.
The black-box discrete optimization (BB-DO) is an important problem that
appears in various real-world problems such as hyper-parameter
optimization of machine learning and is a difficult class of
optimization problems to which optimization methods that require
derivative of an objective function cannot be applied. In addition,
epistasis among parameters, or the dependencies among variables, makes
BB-DO problems more difficult. The bayesian optimization algorithm (BOA)
has been proposed as a promising method to address epistasis in BB-DO
problems. However, BOA suffers from a serious problem. The problem is
that the diversity of a population is likely to be lost. Therefore, BOA
requires a large population size for optimization. In order to remedy
the problem of BOA, we introduce three schemes to maintain the diversity
of the population into BOA. In experiments, we use two benchmark
problems, a 3-deceptive function and a NK-landscape, and a structural
optimization of neural networks to show the effectiveness of the
proposed method. The experimental results showed that the proposed
method improved the number of evaluations by 31.percent and the population
size by 96.percent in a 180-dimensional 3-deceptive function and found
comparable or better solutions in all NK-landscape settings compared to
BOA. In the structural optimization of neural networks, the proposed
method improved the number of evaluations by 21.percent and the population
size by 92.percent compared to BOA. In addition, the proposed method was
superior to conventional optimization methods used in this field,
ASNG-NAS, regularized evolution, reinforcement learning, TPE, and random
search, in terms of the evaluation value.},
keywords = {Machine learning algorithms, Sociology, Neural networks, Optimization methods, Evolutionary computation, Reinforcement learning, Search problems, Black-box discrete optimization, Epistasis among parameters, Bayesian optimization algorithm, Neural architecture search},
doi = {doi:10.1109/CEC45853.2021.9504771},
  notes = {Also known as \cite{9504771}},
)


@INPROCEEDINGS(Unanue:2021:CEC,
%xplor 25 Aug 2021
  author = {Imanol Unanue and Maria Merino and Jose A. Lozano},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A General Framework Based on Walsh Decomposition for
Combinatorial Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {391--398},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper we pursue the use of the Fourier transform for a
general analysis of combinatorial optimization problems. While
combinatorial optimization problems are defined by means of different
notions like weights in a graph, set of numbers, distance between
cities, etc., the Fourier transform allows to put all of them in the
same framework, the Fourier coefficients. This permits its comparison
looking for similarities and differences. Particularly, the Walsh
transform has recently been used over pseudo-boolean functions in order
to design new surrogate models in the black-box scenario and to generate
new algorithms for the linkage discovery problem, among others,
presenting very promising results. In this paper we focus on binary
problems and the Walsh transform. After presenting the Walsh
decomposition and some main properties, we compute the transform of the
Unconstrained Binary Quadratic Problem and several particular cases of
this problem such as the Max-Cut Problem and the Number Partitioning
Problem. The obtained Walsh coefficients not only reinforce the
similarities and differences among the problems which are known in the
literature, but given a set of Walsh coefficients we can say whether or
not they are produced by any of the problems analyzed. Finally, a
geometrical interpretation of the space of Walsh coefficients with
maximum order 2 and the subspace of each analyzed problem is presented.},
keywords = {Couplings, Fourier transforms, Computational modeling, Urban areas, Evolutionary computation, Tools, Optimization, Walsh Functions, Walsh Decomposition, Binary-based Combinatorial Optimization Problems},
doi = {doi:10.1109/CEC45853.2021.9504699},
  notes = {Also known as \cite{9504699}},
)


@INPROCEEDINGS(Pietrantuono:2021:CEC,
%xplor 25 Aug 2021
  author = {Roberto Pietrantuono},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Automated Hypotheses Generation via Combinatorial Causal
Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {399--407},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A powerful form of causal inference employed in many tasks,
such as medical diagnosis, criminology, root cause analysis, biology, is
abduction. Given an effect, it aims at generating a plausible and useful
set of explanatory hypotheses for its causes. This article formulates
the abductive hypotheses generation activity as an optimization problem,
introducing a new class called Combinatorial Causal Optimization
Problems (CCOP). In a CCOP, solutions are in the form of cause-effect
combinations: algorithms are required to construct hypothetical
solutions automatically assessed for plausibility - a mechanism
mimicking the human reasoning when he skims the best solutions from a
set of hypotheses - and for novelty with respect to already known
solutions.The paper presents the CCOP formulation and four real-world
benchmark problems from various domains, released along with artefacts
to implement, run and properly evaluate algorithms for CCOP solutions.
Then, for illustrative purpose, four conventional evolutionary
algorithms are customized to solve CCOPs. Their application demonstrates
the possibility of generating useful solutions (i.e., novel and
realistic hypotheses for a given effect), but also evidences a great
margin for improvement in terms of ratio of good vs bad solutions.},
keywords = {Root cause analysis, Scalability, Evolutionary computation, Benchmark testing, Cognition, Distance measurement, Medical diagnosis, Combinatorial Optimization, Causal inference, Abduction, Learning},
doi = {doi:10.1109/CEC45853.2021.9504816},
  notes = {Also known as \cite{9504816}},
)


@INPROCEEDINGS(Lupoaie:2021:CEC,
%xplor 25 Aug 2021
  author = {Vlad-Ioan Lupoaie and Ivona-Alexandra Chili and Madalina Raschip and Mihaela Elena Breaban},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Evolutionary Approach for Solving Multi-Objective {WCSPs} Using
Mini-Bucket Elimination Heuristics},
year={2021},
  editor = {Yew-Soon Ong},
pages = {408--415},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A Weighted Constraint Satisfaction Problem (WCSP) is a
generalization of a constraint satisfaction problem (CSP) where some of
the constraints can be violated at some cost and thus preferences among
solutions can be expressed. Most research has focused on developing
algorithms for solving single-objective problems. However, many real
world constraint satisfaction problems involve multiple measures of
performance that should be optimized simultaneously. In this paper, we
introduce a new scheme that tackles the Multi-Objective WCSP combining
multi-objective stochastic search and inference methods. The new
approach exploits the efficiency of an evolutionary algorithm and the
advantage of adaptable inference levels offered by the Mini-Bucket
Elimination algorithm. Experiments conducted on the single-objective and
the bi-objective variant of the Weighted Vertex Cover problem
demonstrate the performance of the new approach, highlighting the
benefits the inference step brings to the evolutionary algorithm.},
keywords = {Heuristic algorithms, Computational modeling, Stochastic processes, Evolutionary computation, Inference algorithms, Optimization, multi-objective optimization, weighted constraint satisfaction problems, mini-bucket elimination, evolutionary algorithms, bi-objective weighted vertex cover},
doi = {doi:10.1109/CEC45853.2021.9504973},
  notes = {Also known as \cite{9504973}},
)


@INPROCEEDINGS(Soloviev:2021:CEC,
%xplor 25 Aug 2021
  author = {Vicente P. Soloviev and Concha Bielza and Pedro Larranaga},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Quantum-Inspired Estimation Of Distribution Algorithm To Solve
The Travelling Salesman Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {416--425},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A novel Quantum-Inspired Estimation of Distribution Algorithm
(QIEDA) is proposed to solve the Travelling Salesman Problem (TSP). The
QIEDA uses a modified version of the W state quantum circuits to sample
new solutions during the algorithm runtime. The algorithm behaviour is
compared with other state-of-the-art population-based algorithms. QIEDA
convergence is faster than other algorithms, and the obtained solutions
improve as the size of the problem increases. Moreover, we show that
quantum noise enhances the search of an optimal solution. Because
quantum computers differ from each other, partly due to the topology
that distributes the qubits, the computational cost of executing the
QIEDA in different topologies is analyzed and an ideal topology is
proposed for the TSP solved with the QIEDA.},
keywords = {Computers, Runtime, Qubit, Urban areas, Estimation, Topology, Computational efficiency, quantum computing, quantum machine learning, travelling salesman problem, estimation of distribution algorithm},
doi = {doi:10.1109/CEC45853.2021.9504821},
  notes = {Also known as \cite{9504821}},
)


@INPROCEEDINGS(Santos:2021:CEC,
%xplor 25 Aug 2021
  author = {Jose Santos and Hector Rivas},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolution of Amino Acid Properties in the Context of Protein
Secondary Structure Prediction},
year={2021},
  editor = {Yew-Soon Ong},
pages = {426--433},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Amino acid properties were optimized for protein secondary
structure prediction. The artificial properties were evolved using
differential evolution in a property space of arbitrary dimensionality.
These properties were optimized to provide the correct results in
predicting the elements of the protein secondary structure using a
simple classifier model and with standard benchmark sets. A comparison
is performed with respect to the use of the commonly employed orthogonal
and neutral encoding of the amino acids of the protein chain, together
with a discussion of the similarities of the evolved artificial
properties with respect to physical properties of amino acids.},
keywords = {Proteins, Sequences, Machine learning, Evolutionary computation, Predictive models, Benchmark testing, Amino acids, Differential evolution, Protein structure prediction, Structural biology},
doi = {doi:10.1109/CEC45853.2021.9504939},
  notes = {Also known as \cite{9504939}},
)


@INPROCEEDINGS(Rutkowski:2021:CEC,
%xplor 25 Aug 2021
  author = {Emilia Rutkowski and James Sargant and Sheridan Houghten and Joseph Alexander Brown},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evaluation of Communities from Exploratory Evolutionary
Compression of Weighted Graphs},
year={2021},
  editor = {Yew-Soon Ong},
pages = {434--441},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Contact networks are used as a representation for the modeling
of illness transmission. In this study, we represent not only the links
of the transmissions but also utilize a weighted graph to represent the
probability of transfer. These graphs can be large and complex when
taking into account the number of contacts used in tracing. Compression
of the graph allows for the development of community detection as well
as providing a simpler graph. By examining the contact networks
developed by an evolutionary algorithm for compression, it is discovered
that the choice of fitness function and the appropriate weighting of
edges leads to a different compressed graph, finding different connected
communities; this is also true when compared to the communities
identified by the Louvain community detection algorithm. This
demonstrates the importance of considering weighting in contact
networks, and suggests that in the future an understanding of the
community structure should be utilized by public health officials.},
keywords = {Image edge detection, Evolutionary computation, Public healthcare, Detection algorithms},
doi = {doi:10.1109/CEC45853.2021.9504833},
  notes = {Also known as \cite{9504833}},
)


@INPROCEEDINGS(Barbachan-e-Silva:2021:CEC,
%xplor 25 Aug 2021
  author = {Mariel {Barbachan e Silva} and Pedro {Henrique Narloch} and Marcio Dorn and Pilib O Broin},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Optimisation of Cancer Status Prediction Pipelines using
Bio-Inspired Computing},
year={2021},
  editor = {Yew-Soon Ong},
pages = {442--449},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Cancer is one of the leading causes of death globally, and
early detection is a fundamental factor in improving patient outcomes.
The advent of high-throughput genetic profiling techniques in the last
few decades has led to an explosion of genetic data related to cancer.
Machine learning methods, and classification algorithms in particular,
have been used to find underlying patterns in cancer data and make
diagnostic predictions. The addition of feature selection to
classification pipelines can lead to improvements in predictive
capabilities, since the removal of non-important features benefits the
construction of classification models. We developed a classification
pipeline for cancer status prediction composed of a feature selection
step with SelectKBest and an ensemble classifier system with five
popular supervised learning algorithms. We used three bio-inspired
optimization techniques to select the optimal sets of hyperparameters
for the classification pipeline and compared these approaches on three
cancer microarray datasets. The results indicate that the optimized
pipelines have better predictive performance in all but one of the
experiments compared to the ensemble alone.},
keywords = {Pipelines, Supervised learning, Machine learning, Predictive models, Prediction algorithms, Genetics, Feature extraction, hyperparameter optimization, cancer, machine learning, evolutionary algorithm, swarm intelligence},
doi = {doi:10.1109/CEC45853.2021.9504911},
  notes = {Also known as \cite{9504911}},
)


@INPROCEEDINGS(Jimenez:2021:CEC,
%xplor 25 Aug 2021
  author = {Rodrigo Vega Jimenez and Michael Dube and Sheridan Houghten and James Hughes},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Weighting on the World to Change... an Epidemic},
year={2021},
  editor = {Yew-Soon Ong},
pages = {450--457},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A generative evolutionary algorithm is used to create personal
contact networks representing which individuals can infect others during
an infectious disease scenario. Two problems are considered: (i) finding
networks that maximize the length of a simulated epidemic, and (ii)
finding networks that match given epidemic profiles. A significant
innovation is the introduction of weighted edges to represent the
strength of the contact between individuals. Different weight
initialization conditions are investigated and evaluated for their
performance using a parameter selection mechanism designed to explore
the parameter space. Results show that weighted edges were able to
increase the overall performance achieved by the evolved networks for
both problems considered. Furthermore, it is shown that initializing the
weights with a value greater than one further improves performance. The
results of the parameter selection mechanism were used to test
additional parameter settings thoroughly which further maximize the
length of the simulated epidemic for the evolved graphs.},
keywords = {Epidemics, Technological innovation, Infectious diseases, Evolutionary computation},
doi = {doi:10.1109/CEC45853.2021.9504685},
  notes = {Also known as \cite{9504685}},
)


@INPROCEEDINGS(do-Amaral:2021:CEC,
%xplor 25 Aug 2021
  author = {Laurence Rodrigues {do Amaral} and Alexandre Henrick {da Silva Alves} and Raphael {de Lima Mendes} and Matheus {de Souza Gomes} and Pedro Luiz {Lima Bertarini} and Estevam {Rafael Hruschka}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Applying Never-Ending Learning ({NEL)} Principles to Build a Gene
Ontology (GO) Biocurator},
year={2021},
  editor = {Yew-Soon Ong},
pages = {458--465},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The exponentially growing of generated data and the advanced
computing techniques, catalyzed the necessity of innovative strategies
to store, analyze and capture biological data. The biocuration was
created due to the number and scope of scientific databases in recent
years. Biocuration is an essential part of biological and biomedical
  editor = {Yew-Soon Ong},
research discovery. Extracted from the literature, curated data is so
important to accomplish computational analysis and to train data mining
algorithms. Concurrently, the researchers need a rapidly and more
decisive approach to understand unknown domains. In this paper, we
propose an architecture to help biocurators in Gene Ontology (GO)
classification tasks. Our approach is based on a semi-supervised
environment that couples evolutionary computation (i.e. Genetic
Algorithms (GA)) and traditional classifiers (i.e. Decision Trees and
Naive Bayes). Furthermore, our approach provides high level knowledge
(IF-THEN rules and decision trees), balancing accuracy, interpretability
and comprehensibility that can help GO biocurators in their
classification tasks. Our architecture shows higher classification rates
(about 9percent, i.e., 17,820 correct samples of 18,959), starting from a
small training set, only 2,707 samples (12.percent of 21,666 samples).},
keywords = {Training, Databases, Evolutionary computation, Computer architecture, Ontologies, Biology, Decision trees, Never-Ending Learning, Gene Ontology, Biocuration, Ensemble classifiers, Semi-supervised learning},
doi = {doi:10.1109/CEC45853.2021.9504981},
  notes = {Also known as \cite{9504981}},
)


@INPROCEEDINGS(Fraga:2021:CEC,
%xplor 25 Aug 2021
  author = {Larissa M. Fraga and Gina M. B. {de Oliveira} and Luiz G. A. Martins},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multistage Evolutionary Strategies for Adjusting a Cellular
Automata-based Epidemiological Model},
year={2021},
  editor = {Yew-Soon Ong},
pages = {466--473},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={An epidemiological model based on cellular automata (CA) rules
is tuned through several parameters to provide a more accurate
simulation of the real phenomena. CA are dynamic systems capable of
describing complexity from simple components and local iterations. The
parameters setting discussed here is guided by reference values that
were obtained with real field data. We started from a recent study in
which an adequate parameters configuration was sought for a stochastic
CA-based epidemiological model of Chagas Disease through an evolutionary
approach. The results were satisfactory but the performance of the
standard genetic algorithm (GA) previously employed declines with the
expansion of the search space. In order to improve performance, we
present a multistage evolutionary strategy, where different settings are
applied based on the current stage of the GA search. The proposed
evolutionary approach provided solutions with the least error in the set
of experiments, confirming the improvement over the previous approach.},
keywords = {Graphical models, Computational modeling, Heuristic algorithms, Lattices, Evolutionary computation, Search problems, Genetic algorithms, Multistage evaluation, Genetic algorithm, Cellular automata, Dynamics modeling, Parameters adjustment},
doi = {doi:10.1109/CEC45853.2021.9504738},
  notes = {Also known as \cite{9504738}},
)


@INPROCEEDINGS(Silva:2021:CEC,
%xplor 25 Aug 2021
  author = {Bruno M. P. Silva and Heder S. Bernardino and Helio J. C. Barbosa},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Human Activity Recognition Using Parallel Cartesian Genetic
Programming},
year={2021},
  editor = {Yew-Soon Ong},
pages = {474--481},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Human activity recognition (HAR) is applicable to a wide range
of real-life situations. While machine learning algorithms can be
applied for solving this problem, difficulties remain, such as handling
a large amount of data available for training and selecting the most
appropriate features. Hence, the advent of methods to reduce these
issues and improve the currently available algorithms is relevant. Thus,
we propose here the application of Cartesian Genetic Programming of
Artificial Neural Networks (CGPANN) for training models for HAR. As the
computational cost is a relevant issue in this context, high-performance
computing strategies in graphic processing units (GPU) are proposed for
CGPANN. Two computational experiments are executed and the results show
a decrease in computational time spent with the usage of different data
structures for the parallel CGPANN on the GPU. Moreover, the CGPANN
models for HAR are promising when compared to results from the
literature.},
keywords = {Training, Machine learning algorithms, Computational modeling, Graphics processing units, Genetic programming, Evolutionary computation, Artificial neural networks, Activity Recognition, Cartesian Genetic Programming, Machine Learning, Parallel Computing, GPU},
doi = {doi:10.1109/CEC45853.2021.9504793},
  notes = {Also known as \cite{9504793}},
)


@INPROCEEDINGS(Yuan:2021:CEC,
%xplor 25 Aug 2021
  author = {Yingfang Yuan and Wenjun Wang and Wei Pang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Genetic Algorithm with Tree-structured Mutation for
Hyperparameter Optimisation of Graph Neural Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {482--489},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In recent years, graph neural networks (GNNs) have gained
increasing attention, as they possess the excellent capability of
processing graph-related problems. In practice, hyperparameter
optimisation (HPO) is critical for GNNs to achieve satisfactory results,
but this process is costly because the evaluations of different
hyperparameter settings require excessively training many GNNs. Many
approaches have been proposed for HPO, which aims to identify promising
hyperparameters efficiently. In particular, the genetic algorithm (GA)
for HPO has been explored, which treats GNNs as a black-box model, of
which only the outputs can be observed given a set of hyperparameters.
However, because GNN models are sophisticated and the evaluations of
hyperparameters on GNNs are expensive, GA requires advanced techniques
to balance the exploration and exploitation of the search and make the
optimisation more effective given limited computational resources.
Therefore, we proposed a tree-structured mutation strategy for GA to
alleviate this issue. Meanwhile, we reviewed the recent HPO works, which
gives room for the idea of tree-structure to develop, and we hope our
approach can further improve these HPO methods in the future.},
keywords = {Training, Computational modeling, Evolutionary computation, Graph neural networks, Optimization, Genetic algorithms, generic algorithm, tree-structured mutation, graph neural network, hyperparameter optimisation},
doi = {doi:10.1109/CEC45853.2021.9504717},
  notes = {Also known as \cite{9504717}},
)


@INPROCEEDINGS(Foo:2021:CEC,
%xplor 25 Aug 2021
  author = {Yong Wee Foo and Cindy Goh},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Solar Irradiance Forecasting in Tropical Weather using an
Evolutionary Lean Neural Network},
year={2021},
  editor = {Yew-Soon Ong},
pages = {490--497},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Clean electricity system based on solar energy is rapidly
growing. However, the intermittency of solar power remains an issue. An
accurate solar irradiance forecast can mitigate the impact of variable
solar power generation. In this paper, we present the Evolutionary Lean
Neural Network (EVLNN) for time-series solar irradiance forecasting. The
key novelty of EVLNN lies in incorporating a feedback structure to a
partially connected neural network while using an improved genetic
algorithm to optimize its architecture. We train the model with tropical
weather data to provide a days-ahead forecast of solar irradiance at
four different time-steps: 1-min, 15-min, 30-min, and hourly. We
investigate the effects of fewer input features in model training and
assess the model accuracy using a combination of Root Mean Square Error
(RMSE), Mean Absolute Error (MAE), and Adjusted Coefficient of
Determination (Adjusted R^2 ) metrics. We then compare the results with
those obtained from neural networks modeled using other evolutionary
algorithms (EA), namely the Particle Swarm Optimization, the
Differential Evolution, and the classic Genetic Algorithm. We include
the fully connected nonlinear time-delay backpropagation neural network
as a benchmark against which to evaluate the EA-based models. The
results demonstrated EVLNN's good generalization capability,
specifically in the presence of a sparse dataset. Moreover, our proposed
model trained with a single input variable achieved improved performance
for the hourly, 30-min, and 15-min time-step predictions compared with
those modeled using other learning algorithms.},
keywords = {Training, Computational modeling, Neural networks, Weather forecasting, Solar energy, Predictive models, Solar power generation, evolutionary algorithm, solar irradiance forecasting, solar PV power prediction},
doi = {doi:10.1109/CEC45853.2021.9504875},
  notes = {Also known as \cite{9504875}},
)


@INPROCEEDINGS(Rafe:2021:CEC,
%xplor 25 Aug 2021
  author = {Andrew W. Rafe and Jaime A. Garcia and William L. Raffe},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Exploration Of Encoding And Decoding Methods For Spiking Neural
Networks On The Cart Pole And Lunar Lander Problems Using Evolutionary
Training},
year={2021},
  editor = {Yew-Soon Ong},
pages = {498--505},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Spiking Neural Networks are increasingly drawing interest due
to their potential for large efficiency gains when used with
neuromorphic computers. However, when attempting to replicate the
successes of Artificial Neural Networks, challenges are faced due to
their vastly different architectures and therefore differing methods for
training and optimisation. There has been minimal analysis of the
differences between encoding and decoding methods and the effect of
state space exposure periods on the performance of these networks. The
core contribution of this paper is the detailed analysis of decoding
methods, state exposure periods, and a learned input encoding method of
an evolved Spiking Neural Network within the Reinforcement Learning
context. This is demonstrated using the Cart Pole and Lunar Lander
Reinforcement Learning problems. The paper discovers a negative
correlation between the generation to reach the goal and the state space
exposure period over all decoding methods tested. The state exposure
period is also found to influence the number of random actions taken due
to the decoding methods being unable to select an action. This paper
explores the differences in temporal and rate-based decoding as well as
identifying benefits in resetting networks to their default states
between episode steps. Additionally, the novel input encoder, is
effective at pre-processing state information using the same
evolutionary algorithm as the rest of the network.},
keywords = {Training, Space vehicles, Moon, Data preprocessing, Reinforcement learning, Evolutionary computation, Encoding, genetic algorithm, spiking neurons, spiking neural network, spike train, reinforcement learning},
doi = {doi:10.1109/CEC45853.2021.9504921},
  notes = {Also known as \cite{9504921}},
)


@INPROCEEDINGS(Veal:2021:CEC,
%xplor 25 Aug 2021
  author = {Charlie Veal and Marshall Lindsay and Scott Kovaleski and Derek T. Anderson and Stanton R. Price},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Comparison of Evolutionary and Neural Attention Modeling
Relative to Adversarial Learning},
year={2021},
  editor = {Yew-Soon Ong},
pages = {506--513},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={State-of-the-art machine learning, for computer vision
applications, is based on data-driven feature learning. While these
extraction paradigms often yield impressive results that outperform
human hand-crafted solutions, they unfortunately suffer from a lack of
explainability. In response to this, both the neural network and
evolutionary communities have provided techniques tailored to visually
explain how machines process new observations. Examples range from
gradient-weighted class activation mapping to guided backpropagation,
and from convolutional matrix transpose to improved evolutionary
constructed computing. Previously, we put forth a framework called the
Adversarial Modifier Set (AMS). In AMS, adversarial imagery is generated
based on evolutionary feature identified regions. In this article, we
seek to improve both AMS and general adversarial systems by performing a
comparative analysis between neural attention modeling techniques and
our previously used evolutionary strategy. Preliminary results on a
computer vision dataset show that while the neural techniques are
faster, evolutionary algorithms yield diverse and higher fidelity
attention maps that give rise to improved features for adversarial
learning.},
keywords = {Backpropagation, Computer vision, Analytical models, Computational modeling, Neural networks, Evolutionary computation, Feature extraction, evolutionary algorithms, deep learning, feature identification, attention modeling, adversarial learning},
doi = {doi:10.1109/CEC45853.2021.9504958},
  notes = {Also known as \cite{9504958}},
)


@INPROCEEDINGS(Rios:2021:CEC,
%xplor 25 Aug 2021
  author = {Thiago Rios and Bas {van Stein} and Patricia Wollstadt and Thomas Back and Bernhard Sendhoff and Stefan Menzel},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Exploiting Local Geometric Features in Vehicle Design
Optimization with 3D Point Cloud Autoencoders},
year={2021},
  editor = {Yew-Soon Ong},
pages = {514--521},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Methods for learning and compressing high-dimensional data
allow designers to generate novel and low-dimensional design
representations for shape optimization problems. By using compact design
spaces, global optimization algorithms require less function evaluations
to characterize the problem landscape. Furthermore, data-driven
representations are often domain-agnostic and independent of the user
expertise, and thus potentially capture more relevant design features
than a human designer would suggest. However, more factors than the
dimensionality play a role in the efficiency of design representations.
In this paper, we perform a comparative analysis of design
representations for 3D shape optimization problems obtained with
principal component analysis, kernel-principal component analysis and a
3D point cloud autoencoder, which we apply on a benchmark data set of
computer aided engineering car models. We evaluate the shape-generative
capabilities of these methods and show that we can modify the geometries
more locally with the autoencoder than with the remaining methods. In a
vehicle aerodynamic optimization framework, we verify that this property
of the autoencoder representation improves the optimization performance
by enabling potentially complementary degrees of freedom for the
optimizer. With our study, we provide insights on the qualitative
properties and quantifiable measures on the efficiency of deep neural
networks as shape generative models for engineering optimization
problems, as well as analyses of geometric representations for
engineering optimization with evolutionary algorithms.},
keywords = {Analytical models, Three-dimensional displays, Shape, Drag, Evolutionary computation, Benchmark testing, Aerodynamics, design representation, shape optimization, point cloud autoencoder, sensitivity analysis},
doi = {doi:10.1109/CEC45853.2021.9504746},
  notes = {Also known as \cite{9504746}},
)


@INPROCEEDINGS(Parque:2021:CEC,
%xplor 25 Aug 2021
  author = {Victor Parque},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Differential Particle Scheme with Successful Parent Selection
and its Application to PID Control Tuning},
year={2021},
  editor = {Yew-Soon Ong},
pages = {522--529},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Proportional-integral-derivative (PID) control is ubiquitous
in industrial automation tasks, and the parameter tuning of the gains is
challenging due to nonlinearity and stagnation in local optima. In this
paper we present a differential particle scheme based on
stagnation-based selection mechanism, and evaluate its effectiveness in
the stabilization of a nonlinear inverted pendulum and a magnetic
levitation system. Our computational experiments show the feasibility to
avoid stagnation, the lower variability of convergence over independent
runs, and the feasibility to converge to significantly better fitness
values compared to relevant heuristics in the literature. We believe our
approach offers the building blocks to build stagnation-free nature
inspired optimization algorithms useful for adaptive control and tuning.},
keywords = {PI control, Automation, PD control, Task analysis, Particle swarm optimization, Adaptive control, Tuning, PID Control, Gain Tuning, PID Tuning, Particle Swarm Optimization, Differential Evolution, Stagnation, Optimization},
doi = {doi:10.1109/CEC45853.2021.9504971},
  notes = {Also known as \cite{9504971}},
)


@INPROCEEDINGS(Saemi:2021:CEC,
%xplor 25 Aug 2021
  author = {Behzad Saemi and Mehdi Sadeghilalimi and Ali Asghar {Rahmani Hosseinabadi} and Malek Mouhoub and Samira Sadaoui},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A New Optimization Approach for Task Scheduling Problem Using
Water Cycle Algorithm in Mobile Cloud Computing},
year={2021},
  editor = {Yew-Soon Ong},
pages = {530--539},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Mobile devices are used by numerous applications that
continuously need computing power to grow. Due to limited resources for
complex computing, offloading, a service offered for mobile devices, is
commonly used in cloud computing. In Mobile Cloud Computing (MCC),
offloading decides where to execute the tasks to efficiently maximize
the benefits. Hence, we represent offloading as a Task Scheduling
Problem (TSP). This latter is a Multi-Objective Optimization (MOO)
problem where the goal is to find the best schedule for processing
mobile source tasks, while minimizing both the average processor energy
consumption and the average task processing time. Owing to the
combinatorial nature of the problem, the TSP in MCC is known as NP-hard.
To overcome this difficulty in practice, we adopt meta-heuristic search
techniques as they offer a good trade-off between solution quality and
scalability. More precisely, we introduce a new optimization approach,
that we call Multi-objective Discrete Water Cycle Algorithm (MDWCA), to
schedule tasks from mobile source nodes to processor resources in a
hybrid MCC architecture, including public cloud, cloudlets, and mobile
devices. To evaluate the performance of our proposed approach, we
conducted several comparative experiments on many generated TSP
instances in MCC. The simulation results show that MDWCA outperforms the
state-of-the-art optimization algorithms for several quality metrics.},
keywords = {Cloud computing, Energy consumption, Schedules, Processor scheduling, Computer architecture, Load management, Mobile handsets, Cloud Computing, Mobile Cloud Computing, Task Scheduling, Multi-Objective Optimization, Discrete Optimization, Metaheuristics, Nature-inspired Techniques, Water Cycle Algorithm},
doi = {doi:10.1109/CEC45853.2021.9504780},
  notes = {Also known as \cite{9504780}},
)


@INPROCEEDINGS(Plump:2021:CEC,
%xplor 25 Aug 2021
  author = {Christina Plump and Bernhard J. Berger and Rolf Drechsler},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Domain-driven Correlation-aware Recombination and Mutation
Operators for Complex Real-world Applications},
year={2021},
  editor = {Yew-Soon Ong},
pages = {540--548},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Evolutionary algorithms are a very general method for
optimization problems that allow adaption to many different use cases.
Application to real-world problems usually comes with features as
constraints, dependencies and approximations. When a multidimensional
search space comes with strings attached-- namely dependencies between
its dimensions-- an expression in two ways is possible: Restrictive--as
equalities or inequalities-- or vague--as correlations between dimensions,
for example. Correlations between dimensions are not as easy to grasp as
constraints. Therefore, well-known techniques as death penalty or
penalty functions do not apply directly. We propose new mutation and
recombination operators that incorporate domain knowledge to increase
the offspring fraction that adheres to these correlations. We evaluate
our approach with several benchmark functions and different assumptions
on the dependencies of the search space. We compare the likelihood of
valid (in terms of adhering correlations) outcomes of algorithms using
standard mutation and recombination operators to those with the proposed
operators. We find that the correlation-aware operators preserve
population's features in terms of dependencies.},
keywords = {Correlation, Evolutionary computation, Benchmark testing, Standards, Optimization},
doi = {doi:10.1109/CEC45853.2021.9504931},
  notes = {Also known as \cite{9504931}},
)


@INPROCEEDINGS(Savangouder:2021:CEC,
%xplor 25 Aug 2021
  author = {Ravindra V. Savangouder and Jagdish C. Patra and Suresh Palanisamy},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Hybrid {MLP-PSO-based} Technique to Predict Process Parameters and
Alloying Compositions in ADI for Sustainable Manufacturing},
year={2021},
  editor = {Yew-Soon Ong},
pages = {549--555},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Due to some of the attractive physical properties, e.g., high
ductility and strength, and machinability, austempered ductile iron
(ADI) has been widely used to produce components, in heavy industries,
automobiles, farm machineries, and railways. Depending on the design
requirements, a component requires ADI with specific physical
characteristics. Several alloying elements are added to molten iron and
then it undergoes a complex austempering thermal process to produce ADI.
During the austempering process, the alloying elements interact with
iron in a highly nonlinear complex manner that leads to formation of
microstructure which determines its physical properties. However,
currently, there is no technique available that can specify the
austempering temperature and time, and alloying element proportions to
produce ADI with a specific characteristic. In this paper, we propose a
novel hybrid multilayer perceptron (MLP) and particle swarm optimization
(PSO)-based technique to predict the austempering process parameters and
alloying compositions to produce ADI with a specific physical
characteristic, e.g., Vickers hardness number (VHN). In the first phase,
an MLP is trained to learn the austempering process in a forward
modeling scheme using the experimental data taken from literature. In
the second phase, in an inverse modeling scheme, using the trained MLP
and PSO algorithm, a solution is obtained that provides the predicted
austempered process parameters and alloying elements to produce ADI with
a specific VHN. With extensive simulation results it is shown that the
proposed technique can provide feasible and accurate solutions that
provide optimum use of expensive alloying materials leading to
sustainable manufacturing.},
keywords = {Temperature, Inverse problems, Simulation, Prediction algorithms, Alloying, Iron, Silicon, Particle swarm optimization, multilayer perceptron, inverse modeling, austempered ductile iron, prediction of VHN, sustainable manufacturing},
doi = {doi:10.1109/CEC45853.2021.9505007},
  notes = {Also known as \cite{9505007}},
)


@INPROCEEDINGS(Attaoui:2021:CEC,
%xplor 25 Aug 2021
  author = {Mohammed Oualid Attaoui and Hanene Azzag and Nabil Keskes and Mustapha Lebbah},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A New Subspace Multi-Objective Approach for the Clustering and
Selection of Regions of Interests in Histopathological Images},
year={2021},
  editor = {Yew-Soon Ong},
pages = {556--563},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Histopathology images represent a source of assistance for
pathologists when diagnosing Cancer. However, in histopathology or in
cancer image analysis, pathologists mostly diagnose the pathology as
positive if a small part of it is considered cancer tissue. These small
parts are called regions of interest (ROI) or patches. Finding the
relevant patches is crucial as it can save computation time and memory.
Subspace clustering discovers clusters embedded in multiple, overlapping
subspaces of high dimensional data. It is an extension of feature
selection, which tries to identify relevant subsets of features that are
relevant to the clustering process. However, subspace clustering
algorithms provide a partition of the data based on one cluster validity
measure, assuming a homogeneous similarity measure over the entire data
set makes the algorithms not robust to variations in the data
characteristics. Therefore, it is beneficial to optimize multiple
validity indices simultaneously to capture different aspects of the
datasets. The goal of Multi-Objective clustering methods (MOC) is to
derive significant clusters by applying two or more objective functions.
This paper proposes a new clustering algorithm for patch selection based
on subspace and multi-objective clustering to find the data's best
partitioning and the images' most relevant patches.},
keywords = {Image analysis, Histopathology, Statistical analysis, Clustering algorithms, Evolutionary computation, Linear programming, Feature extraction, Subspace clustering, Multi-objective clustering, Histopathological images, Patch selection},
doi = {doi:10.1109/CEC45853.2021.9504811},
  notes = {Also known as \cite{9504811}},
)


@INPROCEEDINGS(Liao:2021:CEC,
%xplor 25 Aug 2021
  author = {Lushen Liao and Adam Kotaro Pindur and Hitoshi Iba},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Programming with Random Binary Decomposition for
Multi-Class Classification Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {564--571},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper introduces a new Genetic Programming (GP) based
classification framework for multiclass classification problems. The
proposed framework uses a binary decomposition-based GP method to
extract new features to enhance the performance of classifiers in the
multiclass classification task. We firstly introduce a random binary
decomposition method that uses a part-vs-part strategy to decompose the
multiclass problems which increase the number of binary problems that
can be decomposed from a multiclass problem. Then the details of
combining GP with this binary decomposition method for feature
extraction are explained. Finally, we compare our method to several
popular ML methods and traditional GP methods in a broad set of
benchmark problems. The outcome shows the performance of classifiers is
enhanced for multi-class classification tasks when combined with this
technique. The effect of applying this framework to different
classifiers and large real-world data set is also explored. The results
suggest the effectiveness and universality of our method.},
keywords = {Genetic programming, Evolutionary computation, Benchmark testing, Feature extraction, Task analysis, genetic programming, multiclass classification, binary decomposition, feature extraction, feature synthesis},
doi = {doi:10.1109/CEC45853.2021.9504967},
  notes = {Also known as \cite{9504967}},
)


@INPROCEEDINGS(Matousek:2021:CEC,
%xplor 25 Aug 2021
  author = {Radomil Matousek and Rene Pierre Lozi and Tomas Hulka},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Stabilization of Higher Periodic Orbits of the Lozi and Henon
Maps using Meta-evolutionary Approaches},
year={2021},
  editor = {Yew-Soon Ong},
pages = {572--579},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper deals with an advanced adjustment of stabilization
sequences for selected discrete chaotic systems by means of
meta-evolutionary approaches. As the representative models of
deterministic chaotic systems, a two dimensional Lozi map and two
dimensional Henon map were used. The novelty of the approach is in an
effective use of a new type of objective function, which is essential
for the whole optimization process of higher periodic orbits as well as
an effective use of advanced metaheuristic optimization methods.
Although the task of stabilizing the Lozi and Henon chaotic systems is
known, its solution presented for periodic orbit four is not trivial.
The task of stabilizing the Lozi chaotic systems for period four is a
new approach. Furthermore, modern meta-heuristics were used for own
design of the external disturbance sequences. The used optimization
methods are a naive grid-based algorithm (NG), a grid-based Nelder-Mead
Algorithm (NM), a Genetic Algorithm (GA) as well as Genetic Programming
(GP). A connection of GP and second level optimization using GA displays
significantly better results than the given stand-alone meta-heuristic
techniques.},
keywords = {Chaos, Simulation, Genetic programming, Optimization methods, Linear programming, Orbits, Trajectory, Chaos control, Evolutionary computation, Lozi map, Henon map, Optimization},
doi = {doi:10.1109/CEC45853.2021.9504798},
  notes = {Also known as \cite{9504798}},
)


@INPROCEEDINGS(Stapleton:2021:CEC,
%xplor 25 Aug 2021
  author = {Fergal Stapleton and Edgar Galvan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Semantic Neighborhood Ordering in Multi-objective Genetic
Programming based on Decomposition},
year={2021},
  editor = {Yew-Soon Ong},
pages = {580--587},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Semantic diversity in Genetic Programming has proved to be
highly beneficial in evolutionary search. We have witnessed a surge in
the number of scientific works in the area, starting first in discrete
spaces and moving then to continuous spaces. The vast majority of these
works, however, have focused their attention on single-objective genetic
programming paradigms, with a few exceptions focusing on Evolutionary
Multi-objective Optimization (EMO). The latter works have used
well-known robust algorithms, including the Non-dominated Sorting
Genetic Algorithm II and the Strength Pareto Evolutionary Algorithm,
both heavily influenced by the notion of Pareto dominance. These
inspiring works led us to make a step forward in EMO by considering
Multi-objective Evolutionary Algorithms Based on Decomposition (MOEA/D).
We show, for the first time, how we can naturally promote semantic
diversity in MOEA/D in Genetic Programming.},
keywords = {Systematics, Semantics, Sociology, Genetic programming, Evolutionary computation, Ions, Surges, Semantics, Genetic Programming, Multi-objective Evolutionary Algorithms Based on Decomposition, Unbalanced data},
doi = {doi:10.1109/CEC45853.2021.9504860},
  notes = {Also known as \cite{9504860}},
)


@INPROCEEDINGS(Yang:2021:CEC,
%xplor 25 Aug 2021
  author = {Yunhan Yang and Bing Xue and Linley Jesson and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Programming for Symbolic Regression: A Study on Fish
Weight Prediction},
year={2021},
  editor = {Yew-Soon Ong},
pages = {588--595},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The fish weight is a very important factor in fisheries
science and management since it explains the growth and living
conditions of fish populations. A power regression model has been
commonly used to explain the relationship between the fish length and
the weight. In this work, Genetic Programming (GP) for symbolic
regression is used to build a new model for predicting the fish weight,
which allows us to include more features into the model to discover any
hidden relationship, and the GP based symbolic regression makes the
model interpretable comparing with other machine learning methods. A
publicly available dataset is taken with four species of fish which
includes more features than just the fish length that is commonly used
in existing models. The proposed GP based symbolic regression method has
been examined on those four species. The results are compared with the
weight prediction baseline methods including Linear Regression, Power
Regression model, k-Nearest Neighbour, Ridge Regression, Decision Tree,
Random Forest, Gradient Boosting, and Multilayer Perceptron. GP performs
better, or at least as good as the baseline methods on the test set.
Furthermore, the generated GP models also can select different features
for different species to improve the prediction performance due to GP's
explicit feature selection ability. Some models are interpretable with
relatively simple expression. The GP method is also able to find models
that are similar to the power regression model, but more features are
included rather than a single length feature to gain improved prediction
performance.},
keywords = {Computational modeling, Sociology, Linear regression, Genetic programming, Predictive models, Multilayer perceptrons, Fish},
doi = {doi:10.1109/CEC45853.2021.9504963},
  notes = {Also known as \cite{9504963}},
)


@INPROCEEDINGS(Maslyaev:2021:CEC,
%xplor 25 Aug 2021
  author = {Mikhail Maslyaev and Alexander Hvatov},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Objective Discovery of {PDE} Systems Using Evolutionary
Approach},
year={2021},
  editor = {Yew-Soon Ong},
pages = {596--603},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Usually, the data-driven methods of the systems of partial
differential equations (PDEs) discovery are limited to the scenarios,
when the result can be manifested as the single vector equation form.
However, this approach restricts the application to the real cases,
where, for example, the form of the external forcing is of interest for
the researcher and can not be described by the component of the vector
equation. In the paper, a multi-objective co-evolution algorithm is
proposed. The single equations within the system and the system itself
are evolved simultaneously to obtain the system. This approach allows
discovering the systems with the form-independent equations. In contrast
to the single vector equation, a component-wise system is more suitable
for expert interpretation and, therefore, for applications. The example
of the two-dimensional Navier-Stokes equation is considered.},
keywords = {Partial differential equations, Process control, Evolutionary computation, Libraries, Mathematical model, Optimization, Tuning, equation discovery, system discovery, partial differential equation, PDE, data-driven methods, multi-objective optimization, MOEA/DD},
doi = {doi:10.1109/CEC45853.2021.9504712},
  notes = {Also known as \cite{9504712}},
)


@INPROCEEDINGS(Al-Helali:2021:CEC,
%xplor 25 Aug 2021
  author = {Baligh Al-Helali and Qi Chen and Bing Xue and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{GP} with a Hybrid Tree-vector Representation for Instance
Selection and Symbolic Regression on Incomplete Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {604--611},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Data incompleteness is a pervasive problem in symbolic
regression, and machine learning in general. Unfortunately, most
symbolic regression methods are only applicable when the given data is
complete. One common approach to handling this situation is data
imputation. It works by estimating missing values based on existing
data. However, which existing data should be used for imputing the
missing values? The answer to this question is important when dealing
with incomplete data. To address this question, this work proposes a
mixed tree-vector representation for genetic programming to perform
instance selection and symbolic regression on incomplete data. In this
representation, each individual has two components: an expression tree
and a bit vector. While the tree component constructs symbolic
regression models, the vector component selects the instances that are
used to impute missing values by the weighted k-nearest neighbour (WKNN)
imputation method. The complete imputed instances are then used to
evaluate the GP-based symbolic regression model. The obtained
experimental results show the applicability of the proposed method on
real-world data sets with different missingness scenarios. When compared
with existing methods, the proposed method not only produces more
effective symbolic regression models but also achieves more efficient
imputations.},
keywords = {Computational modeling, Genetic programming, Machine learning, Evolutionary computation, Regression tree analysis, Symbolic Regression, Genetic Programming, Incomplete Data, Imputation, Instance Selection},
doi = {doi:10.1109/CEC45853.2021.9504767},
  notes = {Also known as \cite{9504767}},
)


@INPROCEEDINGS(Ahrari:2021:CEC,
%xplor 25 Aug 2021
  author = {Ali Ahrari and Saber Elsayed and Ruhul Sarker and Daryl Essam and Carlos A. {Coello Coello}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Modular Analysis and Development of a Genetic Algorithm with
Standardized Representation for Resource-Constrained Project Scheduling},
year={2021},
  editor = {Yew-Soon Ong},
pages = {612--619},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={There has been a considerable amount of research on the
development of metaheuristic methods for resource-constrained project
scheduling problems. Early methods followed the building blocks and even
the formulation of well-understood metaheuristic methods as well as
simple but effective heuristics such as forward-backward improvement. In
contrast, more recent methods employ less familiar, more complex
(hybrid) metaheuristics and non-standard components and formulations.
Although the former may provide better results on standard test
problems, it is not easy to understand how each component has
contributed to improving the results and why a deviation from
well-established formulations, components and methods was necessary.
This research advances our knowledge about the impact of different
strategies and components of customized genetic algorithms (some of
which have been proposed in this study) on the optimization results.
This task is performed by developing a comprehensive genetic algorithm
with several familiar and potentially effective components. A modular
analysis is then performed in which one component is suppressed at a
time, and the resultant performance decline is analyzed. With hindsight
from the modular analysis, a simple method is suggested and the
importance of each component is clarified. Thus, no further
simplification can be performed without compromising efficiency. Our
preliminary results reveal that this customized genetic algorithm
outperforms many existing methods and can compete with the most
successful ones, which, in many cases, are much more complex than our
approach.},
keywords = {Processor scheduling, Evolutionary computation, Scheduling, Task analysis, Standards, Genetic algorithms, Combinatorial optimization, evolutionary algorithm, heuristic, schedule generation scheme, self-adaptation},
doi = {doi:10.1109/CEC45853.2021.9504950},
  notes = {Also known as \cite{9504950}},
)


@INPROCEEDINGS(Niu:2021:CEC,
%xplor 25 Aug 2021
  author = {Mengyi Niu and Ruochen Liu and Handing Wang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Max-Min Ant System based on Decomposition for the Multi-Depot
Cumulative Capacitated Vehicle Routing Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {620--627},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Multi-depot Cumulative Capacited Vehicle Routing Problem
(MDCCVRP) is a relatively new research field in Vehicle Routing Problems
(VRP), which is composed of several traditional VRP variants. This model
is usually applicable to the logistics and transportation problems after
the disaster. A decomposition-based Max-Min ant system (DMMAS) algorithm
is proposed to solve MDCCVRP in this paper. First of all, a new
indicator which measures the nearness between two routes for multi-depot
problems is proposed. The original problem is decomposed into a series
of smaller subproblems, and then, after the optimization phase, a
specific pheromone communication rule between the master problem and
subproblems is adopted to guide the search direction of ants. Finally,
when the search gets into a halt, a mechanism of perturbation is used to
get rid of the local optimality. The algorithm is tested on many
benchmark problems and the experimental results show that our algorithm
can effectively improve the performance in most cases compared to
several state-of-the-art evolutionary algorithms.},
keywords = {Perturbation methods, Heuristic algorithms, Vehicle routing, Transportation, Evolutionary computation, Benchmark testing, Search problems, Multi-depot Cumulative Capacity Vehicle Problem, decomposition, Max-Min ant system, mechanism of perturbation},
doi = {doi:10.1109/CEC45853.2021.9504891},
  notes = {Also known as \cite{9504891}},
)


@INPROCEEDINGS(Ardeh:2021:CEC,
%xplor 25 Aug 2021
  author = {Mazhar Ansari Ardeh and Yi Mei and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Surrogate-Assisted Genetic Programming with Diverse Transfer for
the Uncertain Capacitated Arc Routing Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {628--635},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Uncertain Capacited Arc Routing Problem (UCARP) is an
important routing problem that can model uncertainties of real-world
scenarios. Genetic Programming (GP) is a powerful method for evolving
routing policies for vehicles to enable them make real-time decisions
and handle environmental uncertainties. When facing various problem
domains, knowledge transfer can improve the effectiveness of the GP
training. Previous studies have demonstrated that due to the existence
of duplicated GP individuals in the source domain, the existing transfer
learning methods do not perform satisfactorily for UCARP. To address
this issue, in this work, we propose a method for detecting duplicates
in the source domain and initialising the GP population in the target
domain with phenotypically unique individuals. Additionally, since the
presence of duplicates can limit the number of good GP individuals, we
propose a surrogate-assisted initialisation approach that is able to
generate much more diversely distributed initial individuals in the
target domain. Our experiments demonstrate that our proposed transfer
learning method can significantly improve the effectiveness of GP for
training new UCARP routing policies. Compared with the state-of-the-art
GP with knowledge transfer, the proposed approach can obtain
significantly better solutions on a wide range of UCRP instances, in
terms of both initial and final quality.},
keywords = {Training, Adaptation models, Uncertainty, Transfer learning, Sociology, Genetic programming, Routing},
doi = {doi:10.1109/CEC45853.2021.9504817},
  notes = {Also known as \cite{9504817}},
)


@INPROCEEDINGS(Wang:2021:CEC,
%xplor 25 Aug 2021
  author = {Shaolin Wang and Yi Mei and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multi-Objective Genetic Programming Approach with Self-Adaptive
? Dominance to Uncertain Capacitated Arc Routing Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {636--643},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Uncertain Capacitated Arc Routing Problem (UCARP) has a
variety of real-world applications. Genetic Programming Hyper-heuristic
(GPHH) is considered a promising technique to handle UCARP. Many
scholars have shown the power of GPHH of evolving effective routing
policies. However, the size of the evolved routing policies is ignored.
Typically, smaller routing policies can have better interpretability and
generalisation. Thus, it is necessary to optimise the size along with
the effectiveness. The objective selection bias issue arises as the size
is much easier to be optimised than effectiveness. The Pareto front is
biased to the size gradually during the evolutionary process. To address
this issue, we develop an ? dominance criteria based Multi-Objective GP
with a self-adaptive ? scheme (?MOGP-sa). The basic idea of the ?
dominance criteria is to set tradeoff rates between objectives. For
different instances, the search space can be very different. In this
case, the self-adaptive ? scheme is employed to automatically tuning the
? value during the evolutionary process so that we can identify a valid
? value for different instances. This paper examines the proposed
algorithm in eight different problem instances. The experimental results
showed that ?MOGP-sa could effectively handle the objective selection
bias issue, and evolve much better Pareto front on Hyper-Volume and
Inverted Generational Distance than the current state-of-the-art MOGP
approach for UCARP in terms of effectiveness and size on all instances.
Also, ?MOGP-sa can evolve much smaller routing policies than the
state-of-art single-objective GPHH without sacrificing effectiveness.},
keywords = {Genetic programming, Evolutionary computation, Routing, Tuning, Optimization, Convergence},
doi = {doi:10.1109/CEC45853.2021.9504956},
  notes = {Also known as \cite{9504956}},
)


@INPROCEEDINGS(Masood:2021:CEC,
%xplor 25 Aug 2021
  author = {Atiya Masood and Gang Chen and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Feature Selection for Evolving Many-Objective Job Shop Scheduling
Dispatching Rules with Genetic Programming},
year={2021},
  editor = {Yew-Soon Ong},
pages = {644--651},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={JSS (Job Shop Scheduling) is a significant and challenging
combinatorial optimization issue. Dispatching rules have been
successfully used to determine scheduling decisions in the JSS
challenges. Genetic programming (GP) has been widely used to discover
and develop dispatching rules for various scheduling problems. However,
there has been relatively little research into feature selection in
GP-HH for many-objective JSS. In many conflicting objective contexts,
it's also vital to quantify the contribution of features. This work
presents a new two-stage GP-HH methodology for many-objective JSS with
feature selection for changing rules. The quality of the solutions
(dispatching rules) after incorporating the many-objective algorithm
with feature selection is investigated in this paper. On a
four-objective JSS problem, the suggested algorithm (FS-GP-NSGA-III) is
compared to the standard GP-NSGA-III. The experimental results show that
using GP to pick relevant features improves the algorithm's performance.
Furthermore, the proposed technique generates rules that are minimal in
size and easy to understand.},
keywords = {Job shop scheduling, Processor scheduling, Genetic programming, Evolutionary computation, Feature extraction, Dispatching, Standards, many-objective optimization, feature selection, genetic programming, hyper-heuristic, job shop scheduling},
doi = {doi:10.1109/CEC45853.2021.9504895},
  notes = {Also known as \cite{9504895}},
)


@INPROCEEDINGS(Srivastava:2021:CEC,
%xplor 25 Aug 2021
  author = {Gaurav Srivastava and Alok Singh and Rammohan Mallipeddi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Hybrid Discrete Differential Evolution Approach for the Single
Machine Total Stepwise Tardiness Problem with Release Dates},
year={2021},
  editor = {Yew-Soon Ong},
pages = {652--659},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, a novel hybrid discrete differential evolution
based approach is proposed to address a single machine scheduling
problem where each job has a release date and the tardiness cost of the
job increases stepwise with respect to various due dates. In the
literature, this problem is termed as the single machine total stepwise
tardiness problem with release dates (SMTSTP-R). The objective of the
problem is to find a schedule of jobs which minimizes the total
tardiness cost. The stepwise increase in tardiness cost is more
prevalent in several real life scenario, especially in transportation.
We have used two constructive heuristics and concept of opposition based
solutions to generate initial population. Our proposed approach uses a
series of local searches to further enhance the quality of solutions
obtained by the proposed discrete differential evolution approach. In
order to justify the superiority of proposed approach, various
comparisons are done with the existing approaches available in the
literature. The results of these comparisons validate the superiority of
our approach in comparison to the existing state-of-the-art approaches.},
keywords = {Schedules, NP-hard problem, Perturbation methods, Sociology, Transportation, Evolutionary computation, Benchmark testing, Discrete differential evolution, Discrete optimization, Scheduling, Single machine total stepwise tardiness problem with release dates},
doi = {doi:10.1109/CEC45853.2021.9504914},
  notes = {Also known as \cite{9504914}},
)


@INPROCEEDINGS(Majkutewicz:2021:CEC,
%xplor 25 Aug 2021
  author = {Jan Majkutewicz and Julian Szymanski and Andrzej Sobecki and Higinio Mora and David Gil},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Embedded Representations of Wikipedia Categories},
year={2021},
  editor = {Yew-Soon Ong},
pages = {660--665},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, we present an approach to building neural
representations of the Wikipedia category graph. We test four different
methods and examine the neural embeddings in terms of preservation of
graphs edges, neighborhood coverage in representation space, and their
influence on the results of a task predicting parent of two categories.
The main contribution of this paper is application of neural
representations for improving the structure of Wikipedia categories
graph. We also show that a neural representation based solely on
categories' names can be an alternative to the other representations
build using more complex approaches.},
keywords = {Electronic publishing, Buildings, Encyclopedias, Evolutionary computation, Internet, Task analysis, information retrieval, neural embeddings, Wikipedia, category graph},
doi = {doi:10.1109/CEC45853.2021.9504802},
  notes = {Also known as \cite{9504802}},
)


@INPROCEEDINGS(Polap:2021:CEC,
%xplor 25 Aug 2021
  author = {Dawid Polap and Marcin Wozniak and Jacek Mandziuk},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Meta-heuristic Algorithm As Feature Selector For Convolutional
Neural Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {666--672},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The huge popularity of heuristics contributes not only to the
improvement and modeling of new solutions but also to their adaptation
to selected goals. Recent years have shown the popularity of their use
also in machine learning as a training algorithm or allowing for the
selection of optimal architecture or hyper-parameters. In this paper, we
propose an adaptation of a nature-inspired algorithm for preprocessing
images in a parallel way for obtaining higher classification results.
The proposed idea is based on analyzing images by heuristic
representative which is Red Fox Optimization Algorithm and returning a
specific value. These values are used in deciding to classify the entire
image or trim it to eliminate unnecessary objects. We modeled this
solution and evaluated using the learning transfer method for VOC 2007
dataset. The obtained results were compared on selected classes to show
the advantages of a proposal.},
keywords = {Training, Analytical models, Machine learning algorithms, Heuristic algorithms, Sociology, Clustering algorithms, Classification algorithms, Red Fox Optimization, Heuristics, Key-points search, Feature selection},
doi = {doi:10.1109/CEC45853.2021.9504915},
  notes = {Also known as \cite{9504915}},
)


@INPROCEEDINGS(Sikora:2021:CEC,
%xplor 25 Aug 2021
  author = {Andrzej Sikora and Adam Zielonka and Marcin Wozniak},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Heuristic Optimization Of 18-pulse Rectifier System},
year={2021},
  editor = {Yew-Soon Ong},
pages = {673--680},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The paper proposes a model of regulation for transformer
secondary voltages in order to improve quality of rectified voltage. For
this purpose, a simulation model for an 18-pulse rectifier system was
presented. We have defined optimization task by using proposed parallel
version of the Polar Bear Optimization (PBO) algorithm. For the
developed method simulation results are presented and discussed.
Research show that proposed parallelization enabled efficient
optimization of the large number of parameters and, at the same time,
shortens simulation time.},
keywords = {Simulation, Computational modeling, Rectifiers, Evolutionary computation, Regulation, Voltage control, Task analysis, Polar Bear Optimization, Heuristics, Rectified voltage, Parallel Computing},
doi = {doi:10.1109/CEC45853.2021.9504783},
  notes = {Also known as \cite{9504783}},
)


@INPROCEEDINGS(Pleszczynski:2021:CEC,
%xplor 25 Aug 2021
  author = {Mariusz Pleszczynski and Adam Zielonka and Dawid Polap and Marcin Wozniak and Jacek Mandziuk},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Polar Bear Optimization For Industrial Computed Tomography With
Incomplete Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {681--687},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this article, Polar Bear Optimization Algorithm (PBO) is
parallelized to solve the problem of computed tomography (CT) with
incomplete data. It is vary hard to model correctly 2D and 3D objects by
using CT scanners when information is incomplete. Our approach is to use
PBO to reduce recovery time and simplify specificity of the phenomenon.
Results from our research show that proposed approach is enabling fast
and accurate reconstruction of objects modeled in projection space.},
keywords = {Three-dimensional displays, Computed tomography, Heuristic algorithms, Computational modeling, Artificial bee colony algorithm, Stability analysis, Task analysis, Polar Bear Optimization, Heuristics, Computed Tomography, Incomplete data, Parallel Computing},
doi = {doi:10.1109/CEC45853.2021.9504750},
  notes = {Also known as \cite{9504750}},
)


@INPROCEEDINGS(Andersen:2021:CEC,
%xplor 25 Aug 2021
  author = {Hayden Andersen and Andrew Lensen and Bing Xue},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Programming for Evolving Similarity Functions Tailored to
Clustering Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {688--695},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Clustering is the process of grouping related instances of
unlabelled data into distinct subsets called clusters. While there are
many different clustering methods available, almost all of them use
simple distance-based (dis)similarity functions such as Euclidean
Distance. However, these and most other predefined dissimilarity
functions can be rather inflexible by considering each feature equally
and not properly capturing feature interactions in the data. Genetic
Programming is an evolutionary computation approach that evolves
programs in an iterative process that naturally lends itself to the
evolution of functions. This paper introduces a novel framework to
automatically evolve dissimilarity measures for a provided clustering
dataset and algorithm. The results show that the evolved functions
create clusters exhibiting high measures of cluster quality.},
keywords = {Measurement, Clustering methods, Clustering algorithms, Genetic programming, Evolutionary computation, Euclidean distance, Iterative methods, Clustering, Genetic Programming, Similarity Function, Feature Selection},
doi = {doi:10.1109/CEC45853.2021.9504855},
  notes = {Also known as \cite{9504855}},
)


@INPROCEEDINGS(Lai:2021:CEC,
%xplor 25 Aug 2021
  author = {Daphne Teck Ching Lai and Yuji Sato},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Hybrid Multiobjective Evolutionary Algorithms for Unsupervised
QPSO, BBPSO and Fuzzy clustering},
year={2021},
  editor = {Yew-Soon Ong},
pages = {696--703},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={While there has been many new developments in multiobjective
evolutionary algorithms, they have not been applied or investigated in
clustering problems. In this paper, ten different unsupervised
clustering techniques applying different MOEA (SPEA2, IBEA, MOEA/D and
MOEA/GLU), PSO (QPSO and BBPSO) and Fuzzy approaches are experimented on
ten public datasets. The rationale to apply MOEA is to increase the
exploitation capabilities of clustering techniques to further refine the
cluster solutions found by fuzzy or PSO clustering. The aim is to
investigate in the performance of different types of MOEA applications
in clustering, determining whether MOEA Fuzzy clustering outperform MOEA
PSO variants. Overall, MOEA/D BBPSO was found to produced the best
results. It outperformed MOEA Fuzzy techniques, having tested on
datasets with high number of classes, that are imbalanced and/or
overlapping classes. IBEA Fuzzy clustering was found to produce the
worst results. MOEA/D clustering was found to perform better than other
MOEA techniques. In this work, we showed that MOEA/D BBPSO clustering
produced the best results on challenging datasets. It was able to use
MOEA/D to deepen its exploitation capability while benefiting from the
exploratory ability of BBPSO when clustering challenging datasets.},
keywords = {Refining, Clustering algorithms, Evolutionary computation, Linear programming, fuzzy clustering, swarm intelligence, cluster analysis, multiobjective evolutionary algorithms},
doi = {doi:10.1109/CEC45853.2021.9504968},
  notes = {Also known as \cite{9504968}},
)


@INPROCEEDINGS(Schofield:2021:CEC,
%xplor 25 Aug 2021
  author = {Finn Schofield and Andrew Lensen},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Using Genetic Programming to Find Functional Mappings for {UMAP}
Embeddings},
year={2021},
  editor = {Yew-Soon Ong},
pages = {704--711},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Manifold learning is a widely used technique for reducing the
dimensionality of complex data to make it more understandable and more
efficient to work with. However, current state-of-the-art manifold
learning techniques -- such as Uniform Manifold Approximation and
Projection (UMAP) -- have a critical limitation. They do not provide a
functional mapping from the higher dimensional space to the
lower-dimensional space, instead, they produce only the
lower-dimensional embedding. This means they are "black-boxes" that
cannot be used in domains where explainability is paramount. Recently,
there has been work on using genetic programming to perform manifold
learning with functional mappings (represented by tree/s), however,
these methods are limited in their performance compared to UMAP. To
address this, in this work we propose utilising UMAP to create
functional mappings with genetic programming-based manifold learning. We
compare two different approaches: one that uses the embedding produced
by UMAP as the target for the functional mapping; and the other which
directly optimises the UMAP cost function by using it as the fitness
function. Experimental results reinforce the value of producing a
functional mapping and show promising performance compared to UMAP.
Additionally, we visualise two-dimensional embeddings produced by our
technique compared to UMAP to further analyse the behaviour of each of
the algorithms.},
keywords = {Manifolds, Visualization, Dermatology, Genetic programming, Evolutionary computation, Cost function, Manifold learning, Manifold learning, Genetic Programming, Dimensionality Reduction, Feature Construction, Feature Selection},
doi = {doi:10.1109/CEC45853.2021.9504848},
  notes = {Also known as \cite{9504848}},
)


@INPROCEEDINGS(Alvarado:2021:CEC,
%xplor 25 Aug 2021
  author = {J. Alvarado and J. Manuel Velasco and F. Chavez and J. Ignacio Hidalgo and F. Fernandez {de Vega}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Blood Glucose Prediction Using a Two Phase {TSK} Fuzzy Rule Based
System},
year={2021},
  editor = {Yew-Soon Ong},
pages = {712--719},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Blood glucose management is a difficult task that people with
diabetes usually have to perform by themselves. An accurate and timely
prediction is vital in order to take decisions and recommend corrective
actions to the patient when the future blood glucose value lies outside
of a target range. It is crucial to predict events like hyperglycemia
and hypoglycemia that may cause health damages in the short term and
potential permanent damages in the long term. The aim of this paper is
to describe our research on predicting blood glucose values using a two
phase Takagi-Sugeno-Kang Fuzzy Rule Based System. The first phase is a
learning process where membership functions and rules are optimised by a
genetic algorithm. In the second phase of tuning, we used a genetic
algorithm to perform the selection and optimisation of the rules. To
train our model we used two different scenarios, What-if and Agnostic;
in both of them the inputs are values measured by a continuous
monitoring glucose system as well as previous carbohydrate intake and
insulin injections. In the What-if scenario, assumed future values of
meals and insulin injections are permitted. On the other hand, in the
Agnostic scenario, only information of the past and present events are
available for the prediction. We trained and tested our system with real
data collected from 10 different diabetic patients, producing 30, 60, 90
and 120 minutes predictions with encouraging accuracy results.},
keywords = {Knowledge based systems, Glucose, Diabetes, Takagi-Sugeno model, Insulin, Task analysis, Blood, Diabetes, Glucose prediction, Fuzzy Rule System},
doi = {doi:10.1109/CEC45853.2021.9504992},
  notes = {Also known as \cite{9504992}},
)


@INPROCEEDINGS(Salgotra:2021:CEC,
%xplor 25 Aug 2021
  author = {Rohit Salgotra and Amiram Moshaiov and Thomas Seidelmann and Dominik Fischer and Sanaz Mostaghim},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Optimal Control Policies to Address the Pandemic Health-Economy
Dilemma},
year={2021},
  editor = {Yew-Soon Ong},
pages = {720--727},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Non-pharmaceutical interventions (NPIs) are effective measures
to contain a pandemic. Yet, such control measures commonly have a
negative effect on the economy. Here, we propose a macro-level approach
to support resolving this Health-Economy Dilemma (HED). First, an
extension to the well-known SEIR model is suggested which includes an
economy model. Second, a bi-objective optimization problem is defined to
study optimal control policies in view of the HED problem. Third, four
multi-objective evolutionary algorithms are applied to perform a study
on the health-economy performance trade-offs that are inherent to the
obtained optimal policies. Finally, the results from the applied
algorithms are compared to select a preferred algorithm for future
studies. As expected, for the proposed models and strategies, a clear
conflict between the health and economy performances is found.
Furthermore, the results suggest that the guided usage of NPIs is
preferable as compared to refraining from employing such strategies at
all. This study contributes to pandemic modeling and simulation by
providing a novel concept that elaborates on integrating economic
aspects while exploring the optimal moment to enable NPIs.},
keywords = {Economics, Pandemics, Computational modeling, Optimal control, Evolutionary computation, Coronaviruses, Optimization, SARS-CoV-2, COVID-19, pandemic model, economic model, control policies, multi-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504758},
  notes = {Also known as \cite{9504758}},
)


@INPROCEEDINGS(Bartz-Beielstein:2021:CEC,
%xplor 25 Aug 2021
  author = {Thomas Bartz-Beielstein and Marcel Droscher and Alpar Gur and Alexander Hinterleitner and Tom Lawton and Olaf Mersmann and Dessislava Peeva and Lennard Reese and Nicolas Rehbach and Frederik Rehbach and Amrita Sen and Aleksandr Subbotin and Martin Zaefferer},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Optimization and Adaptation of a Resource Planning Tool for
Hospitals Under Special Consideration of the COVID-19 Pandemic},
year={2021},
  editor = {Yew-Soon Ong},
pages = {728--735},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Hospitals and health-care institutions need to plan the
resources required for handling the increased load, i.e., beds and
ventilators during the COVID-19 pandemic. BaB-Sim.Hospital, an
open-source tool for capacity planning based on discrete event
simulation, was developed over the last year to support doctors,
administrations, health authorities, and crisis teams in Germany. To
obtain reliable results, 29 simulation parameters such as durations and
probabilities must be specified. While reasonable default values were
obtained in detailed discussions with medical professionals, the
parameters have to be regularly and automatically optimized based on
current data.We investigate how a set of parameters that is tailored to
the German health system can be transferred to other regions. Therefore,
we use data from the UK. Our study demonstrates the flexibility of the
discrete event simulation approach. However, transferring the optimal
German parameter settings to the UK situation does not work--parameter
ranges must be modified. The adaptation has been shown to reduce
simulation error by nearly 7percent. The simulation-via-optimization approach
is not restricted to health-care institutions, it is applicable to many
other real-world problems, e.g., the development of new elevator systems
to cover the last mile or simulation of student flow in academic study
periods.},
keywords = {COVID-19, Adaptation models, Ventilators, Pandemics, Hospitals, Tools, Planning, optimization-via-simulation, surrogate-model-based optimization, COVID-19, hospital resource planning, prediction tool, capacity planning},
doi = {doi:10.1109/CEC45853.2021.9504732},
  notes = {Also known as \cite{9504732}},
)


@INPROCEEDINGS(Cervigon:2021:CEC,
%xplor 25 Aug 2021
  author = {Carlos Cervigon and J. Manuel Velasco and Clara Burgos-Simon and Rafael J. Villanueva and J. Ignacio Hidalgo},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Probabilistic Fitting of Glucose Models with Real-Coded Genetic
Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {736--743},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Type 1 Diabetes patients have to control their blood glucose
levels using insulin therapy. Numerous factors (such as carbohydrate
intake, physical activity, time of day, etc.) greatly complicate this
task. In this article we propose a modeling method that will allow us to
make predictions of blood glucose level evolution with a time horizon of
24 hours. This may allow the adjustment of insulin doses in advance and
could help to improve the living conditions of diabetes patients. Our
approach starts from a system of finite difference equations that
characterizes the interaction between insulin and glucose (in the field,
this is known as a minimal model). This model has several parameters
whose values vary widely depending on patient characteristics and time.
Thus, in the first phase of our strategy, We will enrich the patient's
historical data by adding white Gaussian noise, which will allow us to
perform a probabilistic fitting with a 9percent confidence interval. Then,
the model's parameters are adjusted based on the history of each patient
using a genetic algorithm and dividing the day into 12 time intervals.
In the final stage, we will perform a whole-day forecast from an
ensemble of the models fitted in the previous phase. The validity of our
strategy will be tested using the Parkers' error grid analysis. Our
experimental results based on data from real diabetic patients show that
this technique is capable of robust predictions that take into account
all the uncertainty associated with the interaction between insulin and
glucose.},
keywords = {Uncertainty, Fitting, Predictive models, Probabilistic logic, Glucose, Diabetes, Mathematical model, Diabetes, Glucose prediction, genetic algorithms, Evolutionary computation},
doi = {doi:10.1109/CEC45853.2021.9504836},
  notes = {Also known as \cite{9504836}},
)


@INPROCEEDINGS(Dang:2021:CEC,
%xplor 25 Aug 2021
  author = {Truong Dang and Tien Thanh Nguyen and Carlos {Francisco Moreno-Garcia} and Eyad Elyan and John McCall},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Weighted Ensemble of Deep Learning Models based on Comprehensive
Learning Particle Swarm Optimization for Medical Image Segmentation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {744--751},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In recent years, deep learning has rapidly become a method of
choice for segmentation of medical images. Deep neural architectures
such as UNet and FPN have achieved high performances on many medical
datasets. However, medical image analysis algorithms are required to be
reliable, robust, and accurate for clinical applications which can be
difficult to achieve for some single deep learning methods. In this
study, we introduce an ensemble of classifiers for semantic segmentation
of medical images. The ensemble of classifiers here is a set of various
deep learning-based classifiers, aiming to achieve better performance
than using a single classifier. We propose a weighted ensemble method in
which the weighted sum of segmentation outputs by classifiers is used to
choose the final segmentation decision. We use a swarm intelligence
algorithm namely Comprehensive Learning Particle Swarm Optimization to
optimize the combining weights. Dice coefficient, a popular performance
metric for image segmentation, is used as the fitness criteria.
Experiments conducted on some medical datasets of the CAMUS competition
on cardiographic image segmentation show that our method achieves better
results than both the constituent segmentation models and the reported
model of the CAMUS competition.},
keywords = {Deep learning, Measurement, Image segmentation, Image analysis, Semantics, Prediction algorithms, Classification algorithms, image segmentation, deep neural networks, ensemble learning, ensemble method, particle swarm optimization},
doi = {doi:10.1109/CEC45853.2021.9504929},
  notes = {Also known as \cite{9504929}},
)


@INPROCEEDINGS(Shehu:2021:CEC,
%xplor 25 Aug 2021
  author = {Harisu Abdullahi Shehu and Will Browne and Hedwig Eisenbarth},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Particle Swarm Optimization for Feature Selection in Emotion
Categorization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {752--759},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Emotion categorization plays an important role in
understanding human emotions by artificial intelligence systems such as
robots. It is a difficult task as humans express many features, which
vary over time when showing an emotion. Thus, existing classification
techniques are overwhelmed, and the creation of a subset of appropriate
features is needed. Feature selection can be used to improve the
performance of an emotion categorization task by selecting a subset of
features. This removes irrelevant features. Particle swarm optimization
(PSO) is a meta-heuristic algorithm which has demonstrated excellent
performance in feature selection tasks. However, traditional PSO
algorithms often get trapped in local optima as they use their personal
best and global best to determine their search direction, which may lead
to premature convergence. In this paper, we present a time-based PSO
variant by introducing a time-constant into the velocity update function
of the PSO algorithm to avoid premature convergence, particularly in an
emotion video-frame dataset. The method has been incorporated into
binary and continuous PSO, then compared with the two standard versions
on an emotion video-frame (CK+) dataset, as well as on static emotional
datasets (i.e. the JAFFE and NIMH-ChEFS) to ensure that bias has not
been introduced into the algorithm. While the time-based PSO variant
(both binary and the continuous PSO) have achieved non-significantly
higher performance than the standard PSO algorithms on the JAFFE (77.1percent
vs 75.6percent) and NIMH-ChEFS (71.5percent vs 70.5percent) dataset, the performance is
significantly higher on the CK+ (96.1percent vs 94.0percent) dataset.},
keywords = {Support vector machines, Radio frequency, Feature extraction, Classification algorithms, Task analysis, Particle swarm optimization, Artificial intelligence, Emotion categorization, Facial landmarks, Feature selection, Particle swarm optimization (PSO)},
doi = {doi:10.1109/CEC45853.2021.9504986},
  notes = {Also known as \cite{9504986}},
)


@INPROCEEDINGS(Demir:2021:CEC,
%xplor 25 Aug 2021
  author = {Kaan Demir and Bach Hoai Nguyen and Bing Xue and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-objective Multi-label Feature Selection with an Aggregated
Performance Metric and Dominance-based Initialisation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {760--767},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Feature selection (FS) is an essential pre-processing step
that selects small and informative feature subsets. However, most FS
methods focus on single-label classification, where each instance has
one class label. Multi-label classification (MLC) is increasingly more
common as instances possess a set of class labels. It is more
challenging to perform FS for MLC since it needs to consider the
interactions between labels. Furthermore, since the target output is a
set of labels, there are various metrics for evaluating MLC performance,
and each metric assesses the quality of a feature subset on various
aspects. However, certain MLC metrics are shown to possess conflicting
behaviour. To evolve high-quality feature subsets, it is essential to
consider the trade-offs between metrics, making FS for MLC even more
challenging. Evolutionary multi-objective optimisation (EMO) is a great
technique to optimise multiple (potentially) conflicting metrics,
representing different objectives. However, given a large number of
objectives, EMO algorithms usually do not evolve a diverse set of good
feature subsets, especially when there is a limited computational
resource. This paper proposes a novel approach to reducing the number of
objectives, which is expected to maintain or improve the evolved feature
subsets over the original objectives. We also propose a new
decomposition mechanism based on multiple reference points and a novel
initialization mechanism to enhance the quality of the evolved feature
subsets. The experimental results show that the proposed algorithm can
evolve diverse feature subsets with better trade-off between multiple
performance metrics than recent FS algorithms for MLC.},
keywords = {Measurement, Evolutionary computation, Feature extraction, Classification algorithms, Optimization, Multi-label classification, feature selection, multi-objective optimisation, initialisation methods},
doi = {doi:10.1109/CEC45853.2021.9504960},
  notes = {Also known as \cite{9504960}},
)


@INPROCEEDINGS(Luo:2021:CEC,
%xplor 25 Aug 2021
  author = {Juanjuan Luo and Dongqing Zhou and Lingling Jiang and Huadong Ma},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Entropy Driven Multiobjective Particle Swarm Optimization
Algorithm for Feature Selection},
year={2021},
  editor = {Yew-Soon Ong},
pages = {768--775},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Feature selection is an important research field in machine
learning since high-dimensionality is a common characteristic of
real-world data. It has two main objectives, which are to maximize the
classification accuracy while minimizing the number of selected
features. As the two objectives are usually in conflict with each other,
it makes feature selection a multi-objective problem. However, the large
search space and discrete Pareto front makes it not easy for existing
evolutionary multi-objective algorithms. In order to deal with the above
mentioned difficulties in feature selection, an entropy driven
multiobjective particle swarm optimization algorithm is proposed to
remove redundant feature and decrease computational complexity. First,
its basic idea is to model feature selection as a multiobjective
optimization problem by optimizing the number of features and the
classification accuracy in supervised condition simultaneously. Second,
a particle initialization strategy based on information entropy is
designed to improve the quality of initial solutions, and an adaptive
velocity update rule is used to swap between local search and global
search. Besides, a specified discrete nondominated sorting is designed.
These strategies enable the proposed algorithm to gain better
performance on both the quality and size of feature subset. The
experimental results show that the proposed algorithm can maintain or
improve the quality of Pareto fronts evolved by the state-of-the-art
algorithms for feature selection.},
keywords = {Feature extraction, Entropy, Robustness, Classification algorithms, Particle swarm optimization, Task analysis, Information entropy, feature selection, multiobjective optimization, particle swarm optimization},
doi = {doi:10.1109/CEC45853.2021.9504837},
  notes = {Also known as \cite{9504837}},
)


@INPROCEEDINGS(Altarabichi:2021:CEC,
%xplor 25 Aug 2021
  author = {Mohammed Ghaith Altarabichi and Slawomir Nowaczyk and Sepideh Pashami and Peyman Sheikholharam Mashhadi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Surrogate-Assisted Genetic Algorithm for Wrapper Feature Selection},
year={2021},
  editor = {Yew-Soon Ong},
pages = {776--785},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Feature selection is an intractable problem, therefore
practical algorithms often trade off the solution accuracy against the
computation time. In this paper, we propose a novel multi-stage feature
selection framework utilizing multiple levels of approximations, or
surrogates. Such a framework allows for using wrapper approaches in a
much more computationally efficient way, significantly increasing the
quality of feature selection solutions achievable, especially on large
datasets. We design and evaluate a Surrogate-Assisted Genetic Algorithm
(SAGA) which utilizes this concept to guide the evolutionary search
during the early phase of exploration. SAGA only switches to evaluating
the original function at the final exploitation phase.We prove that the
run-time upper bound of SAGA surrogate-assisted stage is at worse equal
to the wrapper GA, and it scales better for induction algorithms of high
order of complexity in number of instances. We demonstrate, using 14
datasets from the UCI ML repository, that in practice SAGA significantly
reduces the computation time compared to a baseline wrapper Genetic
Algorithm (GA), while converging to solutions of significantly higher
accuracy. Our experiments show that SAGA can arrive at near-optimal
solutions three times faster than a wrapper GA, on average. We also
showcase the importance of evolution control approach designed to
prevent surrogates from misleading the evolutionary search towards false
optima.},
keywords = {Upper bound, Evolutionary computation, Feature extraction, Approximation algorithms, Computational efficiency, Complexity theory, Genetic algorithms, Feature selection, Wrapper, Genetic Algorithm, Progressive Sampling, Surrogates, Meta-models, Evolution Control, Optimization},
doi = {doi:10.1109/CEC45853.2021.9504718},
  notes = {Also known as \cite{9504718}},
)


@INPROCEEDINGS(Li:2021:CEC,
%xplor 25 Aug 2021
  author = {An-Da Li and Bing Xue and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Forward Search Inspired Particle Swarm Optimization Algorithm
for Feature Selection in Classification},
year={2021},
  editor = {Yew-Soon Ong},
pages = {786--793},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Particle swarm optimization (PSO) has been widely used for
feature selection (FS) in classification. However, FS is still a
challenging optimization task for PSO when the dimensionality of data is
high. In this paper, we propose a forward search inspired PSO (FSIPSO)
algorithm to build a wrapper-based FS method. In FSIPSO, the search
space dynamically changes during the evolutionary process. Specifically,
we rank the features according to their single-feature classification
performance and divide the search space into several sub-spaces. A
forward search scheme is proposed to sequentially select the sub-spaces.
The selected sub-spaces construct the search space for FSIPSO. With this
scheme, FSIPSO first searches in a small space to quickly find candidate
solutions (feature subsets) with relatively good performance. Then, the
search space expands with the selection of more sub-spaces, and FSIPSO
can further select informative features in the expanded search space.
Moreover, mutation operations are used in FSIPSO to avoid the premature
problem. The experimental results on 8 UCI datasets have shown that
FSIPSO obtains better FS results with less computation time compared
with benchmark PSO-based FS methods. FSIPSO also obtains better
convergence performance than these methods.},
keywords = {Support vector machines, Heuristic algorithms, Feature extraction, Boosting, Classification algorithms, Particle swarm optimization, Task analysis, Particle swarm optimization, feature selection, classification, sequential forward selection},
doi = {doi:10.1109/CEC45853.2021.9504949},
  notes = {Also known as \cite{9504949}},
)


@INPROCEEDINGS(Ghosh:2021:CEC,
%xplor 25 Aug 2021
  author = {Arka Ghosh and Bing Xue and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Binary Differential Evolution based Feature Selection Method with
Mutual Information for Imbalanced Classification Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {794--801},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The feature selection process aims to eliminate redundant
attributes in the data set, thus it leads to improved classification
accuracy. The solution to a feature selection problem is challenging due
to the ever-increasing data volume. This problem gets more complicated
in the case of imbalanced data sets. Most of the traditional feature
selection methods weigh on the majority class when selecting the
informative feature subset, thus the selected features often get a bias
towards the majority class and neglect the significance of the minority
class in the whole process, which results in poor classification
performance in the case of minority class objects. Multiple evolutionary
algorithms based feature selection methods have been introduced in the
past but most of them ignore the class imbalance problem while selecting
the most informative feature subset. In this article, we propose a
binary differential evolution algorithm with Manhattan distance-based
mutation, which employs a joint mutual information maximization based
feature selection criteria along with a novel class distribution-based
weight assignment scheme to tackle the class imbalance problem. In the
experimental studies, we have tested the performance of the proposed
method on well-known data sets using three widely-used performance
metrics (Average Classification Accuracy, F-measure, G-Means). According
to the empirical results, the proposed method performs better than its
contenders in most of the data sets.},
keywords = {Performance evaluation, Redundancy, Evolutionary computation, Feature extraction, Classification algorithms, Mutual information, Feature Selection, Differential Evolution, Class Imbalance, Evolutionary Algorithms},
doi = {doi:10.1109/CEC45853.2021.9504882},
  notes = {Also known as \cite{9504882}},
)


@INPROCEEDINGS(Sha:2021:CEC,
%xplor 25 Aug 2021
  author = {Zhendong Sha and Ting Hu and Yuanzhu Chen},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Feature Selection for Polygenic Risk Scores using Genetic
Algorithm and Network Science},
year={2021},
  editor = {Yew-Soon Ong},
pages = {802--808},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Many human diseases can be attributed to genetic variations in
the genome. Scientists have been identifying genetic variants associated
with disease risks using population-based data. With this knowledge, an
individual's genetic liability to a disease can be estimated using the
polygenic risk score (PRS), calculated based on their genotype profile.
However, selecting the most predictive genetic variants is challenged by
the high dimensionality of genomics data. Typically, hundreds of
thousands of genetic variants are being tested on their association with
a disease risk. Moreover, the effect of a genetic variant on a disease
risk is often influenced by other variants. It is their interactions
that contribute to a disease risk. In this research, we propose a
feature selection method for PRS assessment that is able to search for
combinations of genetic variants using a genetic algorithm and network
science. Our method provides accurate predictive models for PRS
computation, as well as useful insights into the intertwined
relationships among a large number of genetic variants.},
keywords = {Genomics, Evolutionary computation, Predictive models, Feature extraction, Bioinformatics, Diseases, Genetic algorithms, genetic algorithm, feature selection, population genetics, complex diseases, polygenic risk score},
doi = {doi:10.1109/CEC45853.2021.9504993},
  notes = {Also known as \cite{9504993}},
)


@INPROCEEDINGS(Stanovov:2021:CEC,
%xplor 25 Aug 2021
  author = {Vladimir Stanovov and Shakhnaz Akhmedova and Eugene Semenkin},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{NL-SHADE-RSP} Algorithm with Adaptive Archive and Selective
Pressure for CEC 2021 Numerical Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {809--816},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper proposes a variant of the adaptive differential
evolution algorithm, which combines several important concepts,
including non-linear population size reduction, rank-based selective
pressure in the mutation strategy, adaptive archive set usage, as well
as a set of rules to control crossover rate. The developed approach is
applied to solve the CEC 2021 Bound Constrained Single Objective
Optimization Parametrized Benchmark problems. The performed
computational experiments and their statistical analysis show that the
proposed NL-SHADE-RSP algorithm is capable of demonstrating high
efficiency of fining solutions to biased, shifted and rotated functions
compared to other state-of-the-art algorithms, including the winners of
the previous competitions.},
keywords = {Art, Statistical analysis, Sociology, Evolutionary computation, Benchmark testing, Aerospace electronics, Search problems, optimization, differential evolution, selective pressure, CEC 2021},
doi = {doi:10.1109/CEC45853.2021.9504959},
  notes = {Also known as \cite{9504959}},
)


@INPROCEEDINGS(Brest:2021:CEC,
%xplor 25 Aug 2021
  author = {Janez Brest and Mirjam Sepesy Maucec and Borko Boskovic},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Self-adaptive Differential Evolution Algorithm with Population
Size Reduction for Single Objective Bound-Constrained Optimization:
Algorithm j21},
year={2021},
  editor = {Yew-Soon Ong},
pages = {817--824},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, we propose a new algorithm for solving real
parameter single-objective optimization problems that were prepared for
the CEC 2021 Special Session and Competition on Single Objective Bound
Constrained Numerical Optimization. Single-objective optimization
problems are often very complex and computationally expensive. The
presented algorithm, called j21, uses several mechanisms: two
populations, vectors are chosen from both sub-populations in the
mutation operation, crowding in the big population, population size
reduction, etc. We show the experimental results for each benchmark
function for two scenarios of different dimensions and eight
configuration scenarios as required by the organizers of the CEC 2021
Special Session. We also compare the obtained results of j21 in a
scenario with larger dimension and on one selected configuration with
the original DE and j2020 algorithms.},
keywords = {Sociology, Evolutionary computation, Benchmark testing, Statistics, Optimization},
doi = {doi:10.1109/CEC45853.2021.9504782},
  notes = {Also known as \cite{9504782}},
)


@INPROCEEDINGS(Warchulski:2021:CEC,
%xplor 25 Aug 2021
  author = {Eryk Warchulski and Jaroslaw Arabas},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A New Step-Size Adaptation Rule for {CMA-ES} Based on the
Population Midpoint Fitness},
year={2021},
  editor = {Yew-Soon Ong},
pages = {825--831},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, we present the CMA-ES algorithm with a
step-size adaptation rule which is inspired by the 1/5-th success rule.
The method, called PPMF (Previous Population Midpoint Fitness), adjusts
the step-size multiplier ? using the comparison of fitness values of the
current population of points and the midpoint of the previous
population. In the paper we compare the performance of CMA-ES coupled
with PPMF and with two other step-size adaptation rules: Cumulative
Step-size Adaptation (CSA) and Median Success Rule (MSR). For the
comparison we apply a version of the IPOP-CMA-ES strategy and we test
its performance using three benchmark suites: CEC 2013, CEC 2017 and CEC
2021. The results evidence that the efficiency of PPMF is comparable
with CSA and superior to MSR.},
keywords = {Sociology, Evolutionary computation, Benchmark testing, Statistics, continuous optimization, CMA-ES, step-size adaptation, benchmarking},
doi = {doi:10.1109/CEC45853.2021.9504829},
  notes = {Also known as \cite{9504829}},
)


@INPROCEEDINGS(Biswas:2021:CEC,
%xplor 25 Aug 2021
  author = {Subhodip Biswas and Debanjan Saha and Shuvodeep De and Adam D Cobb and Swagatam Das and Brian A Jalaian},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Improving Differential Evolution through Bayesian Hyperparameter
Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {832--840},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We propose a novel Evolutionary Algorithm (EA) based on the
Differential Evolution algorithm for solving global numerical
optimization problem in real-valued continuous parameter space. The
proposed MadDE algorithm leverages the power of the multiple adaptation
strategy with respect to the control parameters and search mechanisms,
and is tested on the benchmark functions taken from the CEC 2021 special
session & competition on single-objective bound-constrained
optimization. Experimental results indicate that MadDE is able to
achieve superior performance on global numerical optimization problems
when compared against state-of-the-art real-parameter optimizers. We
also provide a hyperparameter optimization algorithm SUBHO for improving
the search performance of any EA by finding an optimal set of control
parameters, and demonstrate its efficacy in enhancing MadDE's
performance on the same benchmark. The source code of our implementation
is publicly available at https://github.com/subhodipbiswas/MadDE.},
keywords = {Sociology, Evolutionary computation, Benchmark testing, Aerospace electronics, Linear programming, Bayes methods, Statistics},
doi = {doi:10.1109/CEC45853.2021.9504792},
  notes = {Also known as \cite{9504792}},
)


@INPROCEEDINGS(Mohamed:2021:CEC,
%xplor 25 Aug 2021
  author = {Ali Wagdy Mohamed and Anas A. Hadi and Prachi Agrawal and Karam M. Sallam and Ali Khater Mohamed},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Gaining-Sharing Knowledge Based Algorithm with Adaptive
Parameters Hybrid with IMODE Algorithm for Solving CEC 2021 Benchmark
Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {841--848},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The initiative to introduce new benchmark problems has drawn
attention to the development of new optimization algorithms. Recently, a
set of constrained benchmark problems has been developed as a addition
to CEC benchmark series. This paper proposed a hybrid variant of gaining
sharing knowledge based algorithm with adaptive parameters and improved
multi-operator differential evolution (IMODE) algorithm, called
APGSK-IMODE. It enhanced the performance of recently developed adaptive
gaining sharing knowledge based algorithm. The performance of
APGSK-IMODE has been tested on CEC2021 benchmark problems which contains
10 test functions with dimensions 10 and 20. The results obtained from
the proposed algorithm have been compared with those obtained from the
rival algorithms. The results elaborate the superiority of APGSK-IMODE.
APGSK-IMODE outperforms the competing algorithms with regard to quality
of solution, robustness and convergence.},
keywords = {Knowledge based systems, Evolutionary computation, Benchmark testing, Probability, Robustness, Optimization, Convergence, CEC 2021, GSK algorithm, LSHADE algorithm, Numerical Optimization},
doi = {doi:10.1109/CEC45853.2021.9504814},
  notes = {Also known as \cite{9504814}},
)


@INPROCEEDINGS(Bujok:2021:CEC,
%xplor 25 Aug 2021
  author = {Petr Bujok and Patrik Kolenovsky},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Differential Evolution with Distance-based Mutation-selection
Applied to CEC 2021 Single Objective Numerical Optimisation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {849--856},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A Differential Evolution (DE) algorithm with distance-based
mutation-selection, population size reduction, and an optional external
archive (DEDMNA) is proposed and tested on the CEC 2021 benchmark suite.
The three well-known mutation variants are chosen in combination with
one crossover for this model. The distances of three newly generated
positions are computed to select the most proper position to evaluate.
In the proposed algorithm, an efficient linear population-size reduction
mechanism is applied. Moreover, an archive is employed to store older
effective solutions. The provided results show that the proposed variant
of DEDMNA is able to solve 64 out of 160 optimisation problems.
Moreover, DEDMNA outperforms the efficient adaptive j2020 variant in 102
problems, and it is worse only in 15 problems out of 160. From the
comparison of DEDMNA with five state-of-the-art DE algorithms, the
superiority of DEDMNA is obvious.},
keywords = {Computational modeling, Sociology, Evolutionary computation, Benchmark testing, Statistics, Optimization, Tuning, differential evolution, mutation, Euclidean distance, population size, archive, benchmark set, optimisation, experimental comparison},
doi = {doi:10.1109/CEC45853.2021.9504795},
  notes = {Also known as \cite{9504795}},
)


@INPROCEEDINGS(Kudela:2021:CEC,
%xplor 25 Aug 2021
  author = {Jakub Kudela},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Novel Zigzag-based Benchmark Functions for Bound Constrained
Single Objective Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {857--862},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The development and comparison of new optimization methods in
general, and evolutionary algorithms in particular, rely heavily on
benchmarking. In this paper, the construction of novel zigzag-based
benchmark functions for bound constrained single objective optimization
is presented. The new benchmark functions are non-differentiable, highly
multimodal, and have a built-in parameter that controls the complexity
of the function. To investigate the properties of the new benchmark
functions two of the best algorithms from the CEC'20 Competition on
Single Objective Bound Constrained Optimization, as well as one standard
evolutionary algorithm, were utilized in a computational study. The
results of the study suggest that the new benchmark functions are very
well suited for algorithmic comparison.},
keywords = {Optimization methods, Evolutionary computation, Benchmark testing, Linear programming, Complexity theory, Standards},
doi = {doi:10.1109/CEC45853.2021.9504720},
  notes = {Also known as \cite{9504720}},
)


@INPROCEEDINGS(Biedrzycki:2021:CEC,
%xplor 25 Aug 2021
  author = {Rafal Biedrzycki},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Comparison with State-of-the-Art: Traps and Pitfalls},
year={2021},
  editor = {Yew-Soon Ong},
pages = {863--870},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={When a new metaheuristic is proposed, its results are compared
with the results of the state-of-the-art methods. The results of that
comparison are the outcome of algorithms' implementations, but the
origin, names, and versions of the implementations are usually not
revealed. Instead, only papers that introduced state-of-the-art are
cited. That approach is generally wrong because algorithms are usually
described in articles in a way that explains the idea that is hidden
behind them but omits the technical details. Therefore, developers have
to fill in these details, and they might do so in different ways. The
paper shows that even implementations made by one author who is the
creator of an algorithm give results which differ considerably from one
another. Therefore, for the comparison purpose, the best possible
implementation should be identified and used. To illustrate how details
that are hidden in the code of implementations influence the quality of
the results, sources of quality differences are tracked down for
selected implementations. It was found that sources of the differences
are hidden in auxiliary code and also stem from implementing a different
version of the algorithm which undergoes development. These findings
imply best practice recommendations for researchers, implementation
developers, and authors of the algorithms.},
keywords = {Constraint handling, Publishing, Evolutionary computation, Best practices, Standards, experiments replication, benchmarking, algorithm-implementation gap, CMA-ES},
doi = {doi:10.1109/CEC45853.2021.9504747},
  notes = {Also known as \cite{9504747}},
)


@INPROCEEDINGS(M?s?r:2021:CEC,
%xplor 25 Aug 2021
  author = {Mustafa M?s?r},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Benchmark Set Reduction for Cheap Empirical Algorithmic Studies},
year={2021},
  editor = {Yew-Soon Ong},
pages = {871--877},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The present paper introduces a benchmark set reduction
strategy that can degrade the experimental evaluation cost for the
algorithmic studies. Algorithm design is an iterative development
process within in a test-revise loop. Starting a devised algorithm's
initial version, it needs to be tested for revealing its advantages and
drawbacks. Its shortcomings lead the designers to modify the algorithm.
Then, the modified algorithm is again tested. This two-step pipeline is
repeated until the algorithm meets the expectations such as delivering
the state-of-the-art results on a set of benchmarks for a specific
problem. That recurring testing step can be a burden for whom with
limited computational resources. The mentioned computational cost can
mainly occur due either high per instance runtime cost or having a large
benchmark set. This study focuses on the cases when a target algorithm
needs to be assessed across large benchmark sets. The idea is to
automatically extract problem instance representation through an
instance-algorithm performance data. The derived representation in the
form of latent features is utilized to determine a small yet a
representative subset of a given large instance set. The proposed
strategy is investigated on the Traveling Thief Problem of 9720
instances. The corresponding performance data is collected by the help
of 21 TTP algorithms. The resulting computational analysis showed that
the proposed method is capable of substantially minimizing the benchmark
instance set size.},
keywords = {Runtime, Pipelines, Clustering algorithms, Benchmark testing, Libraries, Iterative algorithms, Generators, benchmark set reduction, empirical analysis, traveling thief problem},
doi = {doi:10.1109/CEC45853.2021.9505012},
  notes = {Also known as \cite{9505012}},
)


@INPROCEEDINGS(Antonov:2021:CEC,
%xplor 25 Aug 2021
  author = {Kirill Antonov and Maxim Buzdalov and Arina Buzdalova and Carola Doerr},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Blending Dynamic Programming with Monte Carlo Simulation for
Bounding the Running Time of Evolutionary Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {878--885},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={With the goal to provide absolute lower bounds for the best
possible running times that can be achieved by (1 + ?)-type search
heuristics on common benchmark problems, we recently suggested a dynamic
programming approach that computes optimal expected running times and
the regret values inferred when deviating from the optimal parameter
choice.Our previous work is restricted to problems for which transition
probabilities between different states can be expressed by relatively
simple mathematical expressions. With the goal to cover broader sets of
problems, we suggest in this work an extension of the dynamic
programming approach to settings in which it may be difficult or
impossible to compute the transition probabilities exactly, but it is
possible to approximate them numerically, up to arbitrary precision, by
Monte Carlo sampling.We apply our hybrid Monte Carlo dynamic programming
approach to a concatenated jump function and demonstrate how the
obtained bounds can be used to gain a deeper understanding into
parameter control schemes.},
keywords = {Monte Carlo methods, Closed-form solutions, Heuristic algorithms, Evolutionary computation, Benchmark testing, Search problems, Dynamic programming},
doi = {doi:10.1109/CEC45853.2021.9504775},
  notes = {Also known as \cite{9504775}},
)


@INPROCEEDINGS(Wang:2021:CEC,
%xplor 25 Aug 2021
  author = {Peng Wang and Guosong Yang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Using Double Well Function as a Benchmark Function for
Optimization Algorithm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {886--892},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The double well function (DWF) is an important theoretical
model originating from quantum physics and has been used to describe the
energy constraint problem in quantum mechanics and structural chemistry.
Although its form may vary, the DWF has two different local minima in
the one-dimensional case, and the number of local minima increases as
its dimension grows. As a multi-stable function, the DWF is assumed to
be a potential candidate for testing the performance of the heuristic
optimization algorithm, which aims to seek the global minimum. To verify
this idea, a typical DWF is employed in this paper, and a mathematical
analysis is presented herein, and its properties as a benchmark function
are discussed in different cases. In addition, we conducted a set of
experiments utilizing a few optimization algorithms, such as the
multi-scale quantum harmonic oscillator algorithm and covariance matrix
adaptation evolution strategy; thus the analysis has been illustrated by
the results of the numerical experiment. Moreover, to guide the design
of an ideal DWF used as a benchmark function in experiment, different
values of the decisive parameters were tested corresponding to our
analysis, and some useful rules were given based on the discussion of
the results.},
keywords = {Chemistry, Heuristic algorithms, Mathematical analysis, Quantum mechanics, Evolutionary computation, Benchmark testing, Harmonic analysis, double well, double well function, benchmark function, optimization algorithm, performance},
doi = {doi:10.1109/CEC45853.2021.9505009},
  notes = {Also known as \cite{9505009}},
)


@INPROCEEDINGS(Zhan:2021:CEC,
%xplor 25 Aug 2021
  author = {Dawei Zhan and Huanlai Xing},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Population Prescreening Strategy for Kriging-Assisted
Evolutionary Computation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {893--900},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Prescreening strategies have been widely used in
surrogate-assisted evolutionary algorithms for screening out poor
solutions. Existing prescreening strategies are designed for
individual-level selection, i.e. they are used to select individuals
from a set of population members. In this work, we propose a population
prescreening strategy based on the multi-point expected improvement
criterion for Kriging-assisted evolutionary algorithms. In each
generation of the proposed algorithm, the evolutionary operators are
used repeatedly to generate a set of candidate populations. Then, these
candidate populations are prescreened by the multi-point expected
improvement criterion and the population with highest multi-point
expected improvement value is selected for the next generation.
Following this, infill criteria are used to select promising solutions
from the selected population for expensive evaluation. Numerical
experiments on eighteen test problems show that the proposed population
prescreening strategy can improve the optimization efficiency of the
Kriging-assisted evolutionary algorithms significantly without
introducing too much additional computational cost.},
keywords = {Computational modeling, Sociology, Evolutionary computation, Computational efficiency, Statistics, Optimization, Next generation networking, expensive optimization, surrogate-assisted evolutionary computation, Kriging model, expected improvement, prescreening strategy},
doi = {doi:10.1109/CEC45853.2021.9504976},
  notes = {Also known as \cite{9504976}},
)


@INPROCEEDINGS(Namura:2021:CEC,
%xplor 25 Aug 2021
  author = {Nobuo Namura},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Surrogate-assisted Reference Vector Adaptation to Various Pareto
Front Shapes for Many-objective Bayesian Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {901--908},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We propose a surrogate-assisted reference vector adaptation
(SRVA) method to solve expensive multi- and many-objective optimization
problems with various Pareto front shapes. SRVA is coupled with a
multi-objective Bayesian optimization (MBO) algorithm using reference
vectors for scalarization of objective functions. The Kriging surrogate
models for MBO is used to estimate the Pareto front shape and generate
adaptive reference vectors uniformly distributed on the estimated Pareto
front. We combine SRVA with expected improvement of penalty-based
boundary intersection as an infill criterion for MBO. The proposed
algorithm is compared with two other MBO algorithms by applying them to
benchmark problems with various Pareto front shapes. Experimental
results show that the proposed algorithm outperforms the other two in
the problems whose objective functions are reasonably approximated by
the Kriging models. SRVA improves diversity of non-dominated solutions
for these problems with continuous, discontinuous, and degenerated
Pareto fronts. Besides, the proposed algorithm obtains much better
solutions from early stages of optimization especially in many-objective
problems.},
keywords = {Adaptation models, Shape, Computational modeling, Evolutionary computation, Benchmark testing, Approximation algorithms, Linear programming, Bayesian optimization, Kriging, Gaussian process, surrogate model, reference vector, penalty-based boundary intersection},
doi = {doi:10.1109/CEC45853.2021.9504917},
  notes = {Also known as \cite{9504917}},
)


@INPROCEEDINGS(Li:2021:CEC,
%xplor 25 Aug 2021
  author = {Chunna Li and Lianbo Yang and Chunlin Gong},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Data-Driven Multi-Objective Evolutionary Algorithm Based on
Combinatorial Parallel Infilling Criterion},
year={2021},
  editor = {Yew-Soon Ong},
pages = {909--916},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Data-driven multi-objective evolutionary algorithm provides an
effective way to solve multi-objective optimization problems with
computationally expensive black-box functions. In this paper, a
data-driven multi-objective evolutionary algorithm based on
combinatorial parallel infilling (DDMOEA-CPI) is proposed. The
DDMOEA-CPI uses the Kriging model in lieu of the real function, and
combines the infilling criteria based on the multi-objective lower
confidence bound (MLCB), as well as the maximum Kriging error of the
Pareto front, in order to guide the population evolution to quickly
obtain the accurate Pareto solutions. Within the criteria, the hyper
volume improvement function is used to select new samples from the
solutions of the MLCB. A set of benchmark tests in 20 dimensions are
taken to validate and evaluate the performance of the DDMOEA-CPI. The
results show that the proposed method behaves well in the convergence
and diversity of Pareto solutions.},
keywords = {Constraint handling, Computational modeling, Sociology, Evolutionary computation, Benchmark testing, Linear programming, Indexes, multi-objective optimization, data-driven, parallel infilling criterion, Kriging model, Pareto solutions},
doi = {doi:10.1109/CEC45853.2021.9504757},
  notes = {Also known as \cite{9504757}},
)


@INPROCEEDINGS(Li:2021:CEC,
%xplor 25 Aug 2021
  author = {Zhenhua Li and Christine A. Shoemaker},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Hyper-Parameter Optimization for Deep Learning by Surrogate-based
Model with Weighted Distance Exploration},
year={2021},
  editor = {Yew-Soon Ong},
pages = {917--925},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={To improve deep neural net hyper-parameter optimization we
develop a deterministic surrogate optimization algorithm as an efficient
alternative to Bayesian optimization. A deterministic Radial Basis
Function (RBF) surrogate model is built to interpolate previously
evaluated points, and this surrogate model is incrementally updated in
each iteration. The stochastic algorithm CMA-ES is used to search the
acquisition function based on the surrogate. The acquisition function at
a point is based on a weighted average of the surrogate at x and the
minimum distance from x to a previously evaluated point. We evaluate the
proposed algorithm RBF-CMA on hyper-parameter optimization tasks for
deep convolutional neural networks on datasets of CIFAR-10, SVHN, and
CIFAR-100. We show that RBF-CMA achieves a promising performance
especially when the search space dimension is high in comparison to
other algorithms including GP-EI, GP-LCB, and SMBO.},
keywords = {Adaptation models, Computational modeling, Neural networks, Optimization methods, Stochastic processes, Robustness, Bayes methods, hyper-parameter optimization, deep neural networks, surrogate optimization, radial basis function},
doi = {doi:10.1109/CEC45853.2021.9504777},
  notes = {Also known as \cite{9504777}},
)


@INPROCEEDINGS(Polonskaia:2021:CEC,
%xplor 25 Aug 2021
  author = {Iana S. Polonskaia and Nikolay O. Nikitin and Ilia Revin and Pavel Vychuzhanin and Anna V. Kalyuzhnaya},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Objective Evolutionary Design of Composite Data-Driven
Models},
year={2021},
  editor = {Yew-Soon Ong},
pages = {926--933},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, a multi-objective approach for the design of
composite data-driven mathematical models is proposed. It allows
automating the identification of graph-based heterogeneous pipelines
that consist of different blocks: machine learning models, data
preprocessing blocks, etc. The implemented approach is based on a
parameter-free genetic algorithm (GA) for model design called
GPComp@Free. It is developed to be part of automated machine learning
solutions and to increase the efficiency of the modeling pipeline
automation. A set of experiments was conducted to verify the correctness
and efficiency of the proposed approach and substantiate the selected
solutions. The experimental results confirm that a multi-objective
approach to the model design allows us to achieve better diversity and
quality of obtained models. The implemented approach is available as a
part of the open-source AutoML framework FEDOT.},
keywords = {Automation, Computational modeling, Pipelines, Data preprocessing, Machine learning, Evolutionary computation, Data models, AutoML, evolutionary algorithms, multi-objective optimization, model design, composite models},
doi = {doi:10.1109/CEC45853.2021.9504773},
  notes = {Also known as \cite{9504773}},
)


@INPROCEEDINGS(Huang:2021:CEC,
%xplor 25 Aug 2021
  author = {Junhao Huang and Bing Xue and Yanan Sun and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Flexible Variable-length Particle Swarm Optimization Approach
to Convolutional Neural Network Architecture Design},
year={2021},
  editor = {Yew-Soon Ong},
pages = {934--941},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The great success of convolutional neural networks (CNNs) in
image classification benefits from the powerful feature learning
abilities of their architectures. However, arbitrarily constructing
these architectures is very costly in terms of manpower and computation
resources. Recently, particle swarm optimization (PSO), as a promising
evolutionary computation (EC) method, has been used to automatically
search for the CNN architectures and achieved encouraging results on
image classification, but these existing methods are not well-designed
and/or do not have good search capabilities. In this paper, a flexible
variable-length PSO algorithm is proposed for automated CNN architecture
design for image classification tasks. Particularly, an improved
encoding scheme is proposed to truly break the fixed-length
representation constraint of PSO and encode the parameters in a more
meaningful way. In addition, a novel velocity and position updating
approach is developed to update variable-length particles. Experiments
on four benchmark datasets are carried out to confirm the superiority of
the proposed algorithm. It is shown that the proposed method leads to
better results comparing to the state-of-the-art algorithms in terms of
classification performance, parameter size and computational complexity.},
keywords = {Computational modeling, Computer architecture, Benchmark testing, Encoding, Classification algorithms, Convolutional neural networks, Particle swarm optimization, Convolutional neural network, particle swarm optimization, neural architecture search, image classification},
doi = {doi:10.1109/CEC45853.2021.9504716},
  notes = {Also known as \cite{9504716}},
)


@INPROCEEDINGS(Viswambaran:2021:CEC,
%xplor 25 Aug 2021
  author = {Ramya Anasseriyil Viswambaran and Gang Chen and Bing Xue and Mohammad Nekooei},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Two-Stage Genetic Algorithm for Designing Long Short Term Memory
(LSTM) Ensembles},
year={2021},
  editor = {Yew-Soon Ong},
pages = {942--949},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Long Short Term Memory (LSTM) is a special kind of Recurrent
Neural Networks popularly used in various applications. However, using a
single LSTM is often not enough to attain reliable performance on
complicated machine learning tasks. This is because LSTM is sensitive to
the specifics of the training data. Ensemble learning is a promising
approach to improve the performance of LSTMs on complicated tasks.
However, it is difficult to design an ensemble of LSTMs. LSTMs that
constitute the ensemble should be both accurate and diverse. This paper
proposes a new two-phase evolutionary algorithm to design ensembles. The
first phase is to evolve best performing LSTMs automatically. A
connection weight inheritance approach is used in the first phase to
improve the effectiveness and efficiency of the evolutionary process.
The second phase is to design ensembles by choosing suitable LSTMs
without fixing the ensemble size in advance. We use bagging to train the
selected LSTMs to build the ensemble to achieve good diversity among the
LSTMs. The proposed approach is evaluated on various classification
tasks. The results show the effectiveness of the proposed approach and
its significant improvement in performance over many state-of-the-art
machine learning models. The results also show the efficiency of the
proposed approach in comparison with the baseline algorithm.},
keywords = {Training, Machine learning algorithms, Training data, Machine learning, Evolutionary computation, Classification algorithms, Reliability, Long Short Term Memory, Recurrent Neural Network, Deep learning, Ensemble},
doi = {doi:10.1109/CEC45853.2021.9504788},
  notes = {Also known as \cite{9504788}},
)


@INPROCEEDINGS(Zhou:2021:CEC,
%xplor 25 Aug 2021
  author = {Xun Zhou and A. K. Qin and Yanan Sun and Kay Chen Tan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Survey of Advances in Evolutionary Neural Architecture Search},
year={2021},
  editor = {Yew-Soon Ong},
pages = {950--957},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Deep neural networks (DNNs) have been frequently and widely
applied for intelligent systems such as object detection, natural
language understanding and speech recognition. Given a specific problem,
we always aim to construct the most suitable DNN to solve it, which
requires choosing the most appropriate model architecture and seeking
the best model parameters values. However, most existing works focus on
model parameters learning under the assumption that the model
architecture can be manually specified as per prior knowledge and/or
trial-and-error experimentation. To overcome this problem, evolutionary
algorithms (EAs) have been widely used to design model architectures
automatically. Further, EAs have been used for neural network
optimization for more than 30 years. Therefore, in this paper, we review
the evolutionary neural architecture search (ENAS) from the view of the
advanced techniques. We hope this work can provide a comprehensive
understanding of EAs' roles for the readers and focus themselves on ENAS.},
keywords = {Deep learning, Neural networks, Natural languages, Computer architecture, Evolutionary computation, Speech recognition, Object detection, Evolutionary Algorithms, Neural Architecture Search, Optimization},
doi = {doi:10.1109/CEC45853.2021.9504890},
  notes = {Also known as \cite{9504890}},
)


@INPROCEEDINGS(Rogers:2021:CEC,
%xplor 25 Aug 2021
  author = {Brendan Rogers and Nasimul Noman and Stephan Chalup and Pablo Moscato},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolutionary Hyperparameter Optimisation for Sentence
Classification},
year={2021},
  editor = {Yew-Soon Ong},
pages = {958--965},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The performance that Deep Neural Networks can achieve on a
specific task is impacted significantly by the hyperparameters selected.
In order to compare the performance of various Deep Neural Network
architectures on sentence classification, an optimised set of
hyperparameters needs to be found for each architecture. In this work we
use a simple Genetic Algorithm to optimise the hyperparameters of three
different architectures and we evaluate their performance on a suite of
sentence classification benchmarks. We found that a single Genetic
Algorithm is capable of optimising a variety of different architectures
and that the evolved configurations found can compete with those chosen
by experts while using fewer overall trainable parameters. The three
architectures tested are a recurrent neural network and two types of
convolutional networks with a large difference in complexity. Of the
three architectures, optimised for sentence classification, a simple
Convolutional Neural Network was the overall best performer consistently
achieving good performance while using very few trainable parameters.},
keywords = {Deep learning, Recurrent neural networks, Computer architecture, Evolutionary computation, Benchmark testing, Network architecture, Convolutional neural networks},
doi = {doi:10.1109/CEC45853.2021.9504719},
  notes = {Also known as \cite{9504719}},
)


@INPROCEEDINGS(Namekawa:2021:CEC,
%xplor 25 Aug 2021
  author = {Shizuma Namekawa and Taro Tezuka},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolutionary Neural Architecture Search by Mutual Information
Analysis},
year={2021},
  editor = {Yew-Soon Ong},
pages = {966--972},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Neural architecture search (NAS) is a promising approach to
fully automated machine learning that applies to arbitrary tasks.
However, existing NAS methods tend to make a complex network that
requires much time to train and predict. Since machine learning
applications are now often run on mobile devices with limited
computational resources, there is a need to obtain small, efficient, yet
effective network models. This paper proposes MIA-NAS (Mutual
Information Analysis Neural Architecture Search), an evolutionary NAS
algorithm that extends Neural Architecture Search by Hill Climbing
(NASH) by Elsken et al.. The proposed method introduces a procedure that
removes layers that are less contributing to making predictions. The
amount of contribution from each layer is quantified by estimating
mutual information they have with true labels. The experiments showed
that MIA-NAS produces smaller networks with comparative performance to
the ones made by NASH. Besides, removing layers based on mutual
information resulted in better performance than randomly selecting
layers to be removed.},
keywords = {Performance evaluation, Architecture, Neural networks, Computer architecture, Machine learning, Prediction algorithms, Real-time systems, neural architecture search, model selection, mutual information, deep learning},
doi = {doi:10.1109/CEC45853.2021.9504845},
  notes = {Also known as \cite{9504845}},
)


@INPROCEEDINGS(Leonteva:2021:CEC,
%xplor 25 Aug 2021
  author = {Anna Ouskova Leonteva and Radia Hamane and Michel Risser and Anne Jeannin-Girardon and Pierre Parrend and Pierre Collet},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={New Evolutionary Method for Studying Physical Properties of
Magneto Caloric Materials},
year={2021},
  editor = {Yew-Soon Ong},
pages = {973--980},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Magnetic refrigeration (MR) is an alternative technology to
conventional vapour compression with a high potential to reduce energy
consumption and greenhouse gases. The working principle of MR is based
on the property of magneto caloric materials (MCMs) to respond to
applied external magnetic field by a variation in their temperature. To
find the inexpensive MCMs with needed physical properties is still an
issue. This paper describes a new method, whose objective is to automate
the characterization process of MCM physical properties. This method is
based on a multi-objective evolutionary optimization to define the
combination of Hamiltonian model free parameters, which corresponds to
needed physical properties of MCM. The advantage of this method is that
it doesn't need first-principles calculations or huge databases. The
description is followed by a thorough validation of the proposed method
for two alloys. The obtained results are in good agreement with the
experimental measurements.},
keywords = {Temperature, Greenhouse effect, Metals, Evolutionary computation, Tools, Numerical models, Magnetic fields, magneto caloric materials, multi-objective evolutionary optimization, industrial optimization problem, real-world application},
doi = {doi:10.1109/CEC45853.2021.9504988},
  notes = {Also known as \cite{9504988}},
)


@INPROCEEDINGS(Sharma:2021:CEC,
%xplor 25 Aug 2021
  author = {Ashish Sharma and Satyasai Jagannath Nanda and Rahul Kumar Vijay},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Binary {NSGA-II} Model for De-clustering Seismicity of Turkey and
Chile},
year={2021},
  editor = {Yew-Soon Ong},
pages = {981--988},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Seismicity de-clustering is the technique to isolate the
earthquake catalog into aftershock-foreshock (clustered) and background
(random) events. These isolated events are widely used in seismology for
hazard assessment and to design the model for future earthquake
predictions. The key challenge in seismic de-clustering is due to
significant overlapping and high correlation between the space-time
domain of aftershock-foreshock and background events. In this
manuscript, a new model is proposed to de-cluster earthquake catalog
based on a binary Non-dominated sorting genetic (B/NSGA)-II algorithm.
In the fundamental version of the popular NSGA-II algorithm, one
apprehension is that crossover and mutation are performed only on
real-valued population. Here binary domain logical crossover and
mutation operators are employed to optimally segregate the seismic
events. The proposed model is tested on thirty year historical
earthquake catalog of Turkey and Chile. Comparative analysis has been
demonstrated with five benchmark de-clustering techniques. The
simulation results demonstrate the potential of the proposed model
efficiently discriminates the aftershocks and background events in the
two catalogs.},
keywords = {Simulation, Earthquakes, Sociology, Seismology, Benchmark testing, Predictive models, Prediction algorithms, Seismicity De-clustering, Binary Multi-objective Optimization, Binary NSGA-II, Coefficient of Variance, m-Morisita Index},
doi = {doi:10.1109/CEC45853.2021.9504964},
  notes = {Also known as \cite{9504964}},
)


@INPROCEEDINGS(Abdullahi:2021:CEC,
%xplor 25 Aug 2021
  author = {Tassallah Abdullahi and Geoff Nitschke},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Predicting Disease Outbreaks with Climate Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {989--996},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The incidence of most diseases varies greatly with seasons,
and global climate change is expected to increase its risk. Predictive
models that automatically capture trends between climate and diseases
are likely to be beneficial in minimizing disease outbreaks. Machine
learning (ML) predictive analytic tools have been popularized across
many health-care applications, however the optimal task performance of
such ML tools largely depends on manual parameter tuning and
calibration. Such manual tuning significantly limits the full potential
of ML methods, especially for high-dimensional and complex task domains,
as typified by real-world health-care application data-sets.
Additionally, the inaccessibility of many health-care data-sets
compounds innate problems of method comparison, predictive accuracy and
the overall advancement of ML based health-care applications. In this
study we investigate the impact of Relevance Estimation and Value
Calibration, an evolutionary parameter optimization method applied to
automate parameter tuning for comparative ML methods (Deep learning and
Support Vector Machines) applied to predict daily diarrhoea cases across
various geographic regions. Data-augmentation is also used to complement
real-world noisy, sparse and incomplete data-sets with synthetic
data-sets for training, validation and testing. Results support the
efficacy of evolutionary parameter optimization and data synthesis to
boost predictive accuracy in the given task, indicating a significant
prediction accuracy boost for the deep-learning models across all
data-sets.},
keywords = {Support vector machines, Training, Tools, Calibration, Convolutional neural networks, Task analysis, Tuning, Predictive Machine Learning, Evolutionary Parameter Optimization, Deep-Learning, Support Vector Machines},
doi = {doi:10.1109/CEC45853.2021.9504740},
  notes = {Also known as \cite{9504740}},
)


@INPROCEEDINGS(Hafsa:2021:CEC,
%xplor 25 Aug 2021
  author = {Mounir Hafsa and Pamela Wattebled and Julie Jacques and Laetitia Jourdan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multi-Objective Evolutionary Approach to Professional Course
Timetabling: A Real-World Case Study},
year={2021},
  editor = {Yew-Soon Ong},
pages = {997--1004},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper describes the concept and implementation of a
multi-objective approach for professional training scheduling problem at
Mandarine Academy. We address the timetabling of courses and trainers
rostering in a highly constrained environment, in which we are required
to optimize 5 different objectives. We describe the problem by providing
a mathematical formulation of both hard/soft constraints and objectives.
We propose the use of Multi-Objective Evolutionary Algorithms (MOEA's)
like NSGA II and NSGA III with custom genetic operators (Mutation and
Crossover) designed specifically for this problem. To evaluate the
effectiveness of our approach, a parameter tuning phase using the i-race
package is performed, followed by a performance comparison using
real-world test instances varying in complexity. Final results shows a
significant improvement in terms of computation time and solutions
diversity with an interesting performance gap between NSGA II and NSGA
III.},
keywords = {Training, Evolutionary computation, Genetics, Complexity theory, Tuning},
doi = {doi:10.1109/CEC45853.2021.9504797},
  notes = {Also known as \cite{9504797}},
)


@INPROCEEDINGS(Li:2021:CEC,
%xplor 25 Aug 2021
  author = {Fuyou Li and Bei Dong and Xiaojun Wu and HongShang Xu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Geometry Assisted Energy Efficient Sweep Coverage Algorithm For
Wireless Sensor Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1005--1011},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={With the rapid development of wireless sensing technology and
the popularity of the internet of things, applications based on wireless
sensor networks have been applied universally in many scenarios. Sweep
coverage is an emerging technique to improve the quality of service by
promoting the energy efficiency of the sensors. To solve the
distance-sensitive-route scheduling problem, which considering the route
scheduling taking advantage of the sensing radius to reduces the cycle
path of the mobile sensors and improves the feedback time of the
network, this article proposes a three-stage geometry assisted
energy-efficient sweep coverage algorithm to determine the route of
mobile sensors with minimal energy consumption. With analysis of the
distribution of the static monitoring targets, initial sensing points
are determined. To minimize the sweep path in a single cycle, we use an
ant colony optimization algorithm to schedule an energy-efficient
routing. In addition, the sensing point of each target is updated by a
local optimization mechanism according to the route information. To
demonstrate the effectiveness of the proposed algorithm, we compare our
proposed algorithm to a wealth of wireless sensor network scenarios. The
simulation results show that our algorithm can significantly reduce the
length of the cycle route, so as to enhance the energy efficiency of the
wireless sensor networks.},
keywords = {Geometry, Wireless communication, Wireless sensor networks, Schedules, Heuristic algorithms, Simulation, Energy efficiency, sweep coverage, wireless sensor networks, particle swarm optimization, energy efficient},
doi = {doi:10.1109/CEC45853.2021.9504736},
  notes = {Also known as \cite{9504736}},
)


@INPROCEEDINGS(Liu:2021:CEC,
%xplor 25 Aug 2021
  author = {Jiansheng Liu and Satoshi Ono and Bilan Zhu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Three-Dimensional Robot Motion Design by Combining Interactive
and Non-Interactive Evolutionary Computation for an Intelligent
Transformable Phone Robot: BaBi},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1012--1019},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper presents an interactive method for designing robot
facial expressions and motions of an intelligent transformable phone
robot BaBi. Designing an objective function to evaluate facial
expressions and motions suitable for BaBi with an original structure is
quite hard although users can easily imagine favorite motions.
Therefore, the proposed method employs Interactive Evolutionary
Computation (IEC) in addition to general optimization. The method allows
users to add candidates to a case base when finding satisfying
candidates and evaluates candidates' scores according to the case base.
It applies a Different Evolution (DE) method to generate new candidates.
Users may alternately push the buttons to generate face and body motion
without making any selection of candidates with a selection interface.
The method will make selection with the scores when users do not make
any selection of candidates. Users can select or not select. It is an
IEC process when users make selections, while it is a non-Interactive
Evolutionary Computation (non-IEC) process when users do not make any
selection, resulting in being a fusion of IEC and non-IEC. Adding more
candidates to the case base can speed up generating the desired motion
and promote the non-IEC progress and reduce the burden on users.
Experimental results have shown that the proposed method produces the
robot motions effectively.},
keywords = {Robot motion, IEC, Evolutionary computation, Linear programming, Robots, Optimization, Faces, Interactive Evolutionary Computation, Non-Interactive Evolutionary Computation, Robot Motion Design, Phone Robot},
doi = {doi:10.1109/CEC45853.2021.9504933},
  notes = {Also known as \cite{9504933}},
)


@INPROCEEDINGS(Malysiak-Mrozek:2021:CEC,
%xplor 25 Aug 2021
  author = {Bozena Malysiak-Mrozek and Lukasz Bozek and Dariusz Mrozek},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Fuzzy Filtering in Large-Scale Prediction of Intrinsically
Disordered Regions of Proteins on Apache Spark},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1020--1027},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Intrinsically disordered proteins (IDPs) participate in many
cellular processes. They are also studied for their participation in the
course and formation of many diseases. Experimental determination of
disordered regions (IDRs) is costly and not always possible. Due to the
exponential growth of protein sequences, for which the 3D structure
cannot be experimentally determined, computational prediction becomes an
important alternative. Spark-IDPP is the large-scale meta-predictor for
IDRs and IDPs designed to run on the Apache Spark cluster. The
meta-prediction with Spark-IDPP includes fuzzy filtering of produced
prediction output. Here, we experimentally validate various fuzzy
filters and show that a properly designed characteristic function for
fuzzy filtering may improve the prediction quality in all modes of the
Spark-IDPP execution.},
keywords = {Proteins, Fuzzy sets, Uncertainty, Three-dimensional displays, Filtering, Cluster computing, Tools, bioinformatics, proteins, fuzzy sets, intrinsically disordered proteins, Apache Spark, computational intelligence},
doi = {doi:10.1109/CEC45853.2021.9504834},
  notes = {Also known as \cite{9504834}},
)


@INPROCEEDINGS(Zorarpac?:2021:CEC,
%xplor 25 Aug 2021
  author = {Ezgi Zorarpac?},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Hybrid Dimension Reduction Based Linear Discriminant Analysis
for Classification of High-Dimensional Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1028--1036},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Linear discriminant analysis (LDA) is a notable classification
algorithm thanks to its major success in many applications of the
real-world. In spite of its successfulness for low-dimensional data, a
dimension reduction is inevitable for its achievement with
high-dimensional data, especially in which the number of features is
more than the training sample size or close to the training sample size.
Principal component analysis (PCA) plus LDA (PCA+LDA), a quite popular
technique, is widely used for raising the classification performance of
LDA over high-dimensional data. However, PCA ignores the label
information in the data. On the other hand, the reduced dimensional data
through PCA still includes indiscriminate (i.e., irrelevant) features.
To cope with the dimensionality problem of LDA, a hybrid dimension
reduction approach of supervised and unsupervised algorithms is proposed
in this study. In the supervised part of the proposed hybrid dimension
reduction method, called DBDERF+PCA, we propose to combine an ensemble
classifier (i.e., random forest) with dichotomous binary differential
evolution (DBDE), a recently proposed variant of binary differential
evolution, by introducing a robust wrapper feature selection. On the
other hand, unsupervised part of the proposed hybrid dimension reduction
method utilizes PCA. The experimental results show that DBDERF+PCA
outperforms PCA in terms of dimension reduction. Thereupon, this hybrid
dimension reduction based LDA, called DBDERF+PCA+LDA, performs better
than PCA+LDA and LDA in terms of run-time and classification
performances.},
keywords = {Dimensionality reduction, Training, Sonar, Lung cancer, Artificial neural networks, Learning (artificial intelligence), Classification algorithms, dimension reduction, dichotomous binary differential evolution, linear discriminant analysis, supervised learning, unsupervised learning},
doi = {doi:10.1109/CEC45853.2021.9504951},
  notes = {Also known as \cite{9504951}},
)


@INPROCEEDINGS(Elewah:2021:CEC,
%xplor 25 Aug 2021
  author = {Abdelrahman Elewah and Abeer A. Badawi and Haytham Khalil and Shahryar Rahnamayan and Khalid Elgazzar},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={3D-{RadViz:} Three Dimensional Radial Visualization for Large-Scale
Data Visualization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1037--1046},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper presents 3D-RadViz, a visualization method for high
dimensional data using a three dimensional radial visualization
technique. The proposed technique extends the capabilities of the
classical two dimensional radial visualization (RadViz) in order to
reduce overlapping data points. For evaluation purposes, the paper
applies the proposed 3D-RadViz alongside with t-SNE and a recently
published 3-dimensional radial visualization technique to the same
real-world datasets from the UCI machine learning repository. The
contribution of the work is an interactive, deterministic, and
high-performance library, implemented in Python, that can be utilized to
realize better high dimensional data visualization using the proposed
3D-RadViz technique. The paper also suggests some directions for future
enhancements to the proposed visualization approach.},
keywords = {Visualization, Filtration, Data visualization, Lattices, Machine learning, Evolutionary computation, Benchmark testing, Data visualization, two dimensional radial visualization (RadViz), three dimensional radial visualization (3D-RadViz), Magic Ball, RadViz-Plotly, large-scale visualization, high dimensional visualization},
doi = {doi:10.1109/CEC45853.2021.9504983},
  notes = {Also known as \cite{9504983}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {An Chen and Zhigang Ren and Yongsheng Liang and Daofu Guo},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Enhancing Cooperative Coevolution for Large Scale Optimization by
Exploiting Decomposition Solutions},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1047--1053},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Problem decomposition plays a vital role in ensuring the
performance of cooperative coevolution (CC) for large scale global
optimization (LSGO). However, existing high-precision decomposition
algorithms require evaluating numerous solutions generally generated in
a fixed style and directly abandon them after the decomposition process.
This leads to the waste of many computation resources. Directing against
this issue, this study attempts to sample decomposition solutions in a
local-search manner such that these solutions not only satisfy the
decomposition requirements but also possess high quality, and thus can
provide a good initial population for the subsequent optimization
process. Following this research idea, this study develops a new
decomposition algorithm named recursive differential grouping with local
search ability (LS-RDG) by embedding the Solis Wets local search
operator into the recently developed RDG algorithm. LS-RDG can obtain
more promising solutions without consuming extra fitness evaluations.
Comprehensive experimental results on two widely used LSGO benchmark
suites verify the effectiveness of LS-RDG and indicate that LS-RDG can
help CC achieve better optimization performance.},
keywords = {Sociology, Evolutionary computation, Benchmark testing, Iron, Statistics, Optimization, cooperative coevolution, large scale global optimization, decomposition algorithm, local search},
doi = {doi:10.1109/CEC45853.2021.9504977},
  notes = {Also known as \cite{9504977}},
)


@INPROCEEDINGS(Maliszewski:2021:CEC,
%xplor 25 Aug 2021
  author = {Michal Maliszewski and Urszula Boryczka},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Using {MajorClust} Algorithm for Sandbox-based {ATM} Security},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1054--1061},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Automated teller machines are affected by two kinds of
attacks: physical and logical. It is common for most banks to look for
zero-day protection for their devices. The most secure solutions
available are based on complex security policies that are extremely hard
to configure. The goal of this article is to present a concept of using
the modified MajorClust algorithm for generating a sandbox-based
security policy based on ATM usage data. The results obtained from the
research prove the effectiveness of the used techniques and confirm that
it is possible to create a division into sandboxes in an automated way.},
keywords = {Whitelists, Online banking, Filtering, Software algorithms, Evolutionary computation, Tools, Big Data, ATM, Clustering, Cyber Security, Whitelisting, Intrusion Protection, Sandboxing, Software Security, Evolutionary Computing, Big Data},
doi = {doi:10.1109/CEC45853.2021.9504862},
  notes = {Also known as \cite{9504862}},
)


@INPROCEEDINGS(Bolufe-Rohler:2021:CEC,
%xplor 25 Aug 2021
  author = {Antonio Bolufe-Rohler and Dania Tamayo-Vera},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Exploration-only Exploitation-only Hybrid for Large Scale
Global Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1062--1069},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Two factors affect the effectiveness of exploration, the bias
introduced by selection and the concurrence of exploration and
exploitation. The Leaders and Followers metaheuristic was designed to
reduce the bias from selection by using a two-population scheme. Minimum
Population Search was designed to limit the concurrence of exploration
and exploitation through the use of Thresheld Convergence in its
sampling strategy. This paper presents Unbiased Exploratory Search,
which combines both approaches and simultaneously addresses the effects
of these two factors. An exploration-only exploitation-only hybrid is
then presented using Unbiased Exploratory Search for the
exploration-only phase of the hybrid. The hybrid is tested on the CEC
large scale optimization benchmark.},
keywords = {Machine learning algorithms, Limiting, Sociology, Machine learning, Evolutionary computation, Benchmark testing, Statistics, unbiased exploratory search, large scale global optimization, hybrid heuristic, exploration},
doi = {doi:10.1109/CEC45853.2021.9504812},
  notes = {Also known as \cite{9504812}},
)


@INPROCEEDINGS(Marchi:2021:CEC,
%xplor 25 Aug 2021
  author = {Felipe Marchi and Rafael Stubs Parpinelli},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multi-objective Approach to the Protein Structure Prediction
Problem using the Biased Random-Key Genetic Algorithm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1070--1077},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Proteins are base molecules present in live organisms. The
study of their structures and functions is of considerable importance
for many application fields, particularly for the pharmaceutical area.
This paper presents a multi-objective model to the Protein Structure
Prediction problem, using three objectives: energy score, secondary
structure information, and contact maps information. A BRKGA method is
used as a global optimizer and adapted to work with multiple objective
problems. Also, the MUFOLD-CL clustering method is applied to select a
single predicted structure. Experiments were carried out to analyze the
proposed model performance using state-of-the-art ab initio algorithms
for comparison. Results obtained indicate that the proposed approach is
competitive in terms of RMSD and GDT metrics.},
keywords = {Proteins, Measurement, Adaptation models, Evolutionary computation, Predictive models, Prediction algorithms, Organisms, structural bioinformatics, evolutionary algorithms, continuous optimization, multiple objectives},
doi = {doi:10.1109/CEC45853.2021.9504745},
  notes = {Also known as \cite{9504745}},
)


@INPROCEEDINGS(Engholdt:2021:CEC,
%xplor 25 Aug 2021
  author = {Kaelan Engholdt and H. David Mathias},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Biologically-Inspired Model for Mass Extinction in Genetic
Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1078--1085},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Mass extinction events have previously been shown to be a
catalyst for accelerating the rate of evolution in genetic algorithms.
This increased evolution rate combined with a destabilization of the
dynamic equilibrium of the genetic algorithm can allow the algorithm to
overcome local maxima in the solution space; however, most
implementations of mass extinction squander the full benefit of the
increased evolutionary rate by relying upon random chance to generate
better solutions in the post-extinction generations. This paper examines
the hypothesis that, by increasing variability and decreasing selection
pressure immediately after an extinction event, the algorithm has a
higher chance of yielding more fit individuals and overcoming local
maxima in the post-extinction generations. We compare the performance of
a genetic algorithm without extinction with the two approaches described
above on a problem with a large search space containing many local
optima, point scattering in the unit circle, and examine the effects of
extinction methods on performance. The preliminary results presented
here show that decreasing selection pressure and increasing variability
help the genetic algorithm achieve higher success rates than with
standard mass extinction.},
keywords = {Evolution (biology), Biological system modeling, Heuristic algorithms, Scattering, Evolutionary computation, Search problems, Path planning},
doi = {doi:10.1109/CEC45853.2021.9504851},
  notes = {Also known as \cite{9504851}},
)


@INPROCEEDINGS(Alves:2021:CEC,
%xplor 25 Aug 2021
  author = {Alexandre Henrick da Silva Alves and Raphael de Lima Mendes and Matheus de Souza Gomes and Pedro Luiz Lima Bertarini and Laurence Rodrigues {do Amaral}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{IG-CEE:} An Embedded Information Gain approach to Genetic
Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1086--1092},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The classification task is among the most used in Data Mining
and it is widely researched nowadays. Some works have already been
developed using Genetic Algorithms (GAs) for classification tasks,
building simple and small IF-THEN classification rules with good
classification results. These methods produce understandable outputs
(IF-THEN rules), unlike some of the traditional classifiers, e.g. SVM
and Artificial Neural Networks, known as black-box type. In this paper,
we proposed a new method called Information Gain-based Computational
Evolutionary Environment (IG-CEE). The proposed method IG-CEE extends
the GA proposed by Amaral and Hruschka named Computational Evolutionary
Environment (CEE). The IG-CEE uses the Information Gain approach to
select a better subset of attributes to compose the IF-THEN rules,
improving classification accuracy and convergence rates. The IG-CEE was
compared with CEE and four (4) traditional classifiers: J48, IBK, Naive
Bayes, and SVM. To compare all methods, we built three (3) synthetic
datasets, generated by GeNIe software. To realize a fair comparison, all
methods were executed, with distinct random seeds, 100 times. For each
method, the final result was calculated using a confidence interval with
9percent of confidence. The proposed method IG-CEE showed better
classification rates in several comparisons.},
keywords = {Support vector machines, Buildings, Evolutionary computation, Artificial neural networks, Software, Data mining, Task analysis, genetic algorithms, classification task, information gain, feature selection, IF-THEN rules, high dimensional data},
doi = {doi:10.1109/CEC45853.2021.9504776},
  notes = {Also known as \cite{9504776}},
)


@INPROCEEDINGS(Jimenez-Aparicio:2021:CEC,
%xplor 25 Aug 2021
  author = {Aaron Jimenez-Aparicio and Efren Mezura-Montes and Hector-Gabriel Acosta-Mesa},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolutionary Algorithms for Searching Almost-Equienergetic Graphs},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1093--1098},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Many real-world problems can be modeled using graphs. This
representation allows us to understand some important aspects of problem
behavior through the calculation of different measures. The energy of a
graph is an invariant measure that has gained interest in network
analysis due to its recent applications, then it is interesting to know
the structural properties that make two graphs with similar energies.In
this work we propose the search for almost-equienergetic graphs using
evolutionary algorithms for Erdos-Renyi networks and trees. The proposed
evolutionary algorithms are able to obtain almost-equienergetic graphs
in larger instances with respect to traditional methods and the results
lead to analyze interesting properties of the graphs found.},
keywords = {Energy measurement, Evolutionary computation, Gain measurement, Almost-equienergetic, Graph energy, Evolutionary algorithm, Graph theory},
doi = {doi:10.1109/CEC45853.2021.9504751},
  notes = {Also known as \cite{9504751}},
)


@INPROCEEDINGS(Nomura:2021:CEC,
%xplor 25 Aug 2021
  author = {Masahiro Nomura and Nobuyuki Sakai and Nobusumi Fukushima and Isao Ono},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Distance-weighted Exponential Natural Evolution Strategy for
Implicitly Constrained Black-Box Function Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1099--1106},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper presents a natural evolution strategy for
implicitly constrained black-box function optimization. The black-box
function optimization is challenging because explicit representations of
objective functions are not given, and only evaluation values of
solutions can be used. In implicitly constrained black-box function
problems, constraints are not explicitly given, and only the feasibility
of a solution is obtained when the objective function is evaluated. In
other words, the amount of constraint violation cannot be obtained,
which makes the optimization difficult. Natural Evolution Strategies
(NES) is one of the promising frameworks for black-box function
optimization. DX-NES is an improved version of xNES which is a promising
NES using a multivariate normal distribution as the probability
distribution. DX-NES has been reported to show good performance on
unconstrained black-box function optimization problems. However, DX-NES
has a serious problem in that its performance degrades when applied to
implicitly constrained problems. In order to address the problem, we
propose DX-NES taking account of Implicit Constraint (DX-NES-IC). In
experiments using benchmark problems and a lens system design problem,
DX-NES-IC showed better performance than DX-NES, xNES, CMA-ES, and those
with the resampling technique in terms of the number of successful
trials and that of evaluations, where the resampling technique is a
constraint handling method which can be used for implicitly constrained
problems.},
keywords = {Estimation, Evolutionary computation, Gaussian distribution, Benchmark testing, Linear programming, Probability distribution, Optimization, Natural Evolution Strategy, Implicit Constraint},
doi = {doi:10.1109/CEC45853.2021.9504865},
  notes = {Also known as \cite{9504865}},
)


@INPROCEEDINGS(Nunes:2021:CEC,
%xplor 25 Aug 2021
  author = {Wesley H. B. Nunes and Mayron C. O. Moreira and Marina Andretta},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Genetic Algorithm for the Nesting Problem With Continuous
Rotations},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1107--1114},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Nesting problem, or the irregular cutting and packing
problem, aims to find the best position of irregular pieces within a
board, minimizing the space used by them. The problem's relevance is
that it is widely used in the furniture, textile, and footwear
industries. We consider a two-dimensional scope with convex and
non-convex parts, with continuous rotation. We implement a genetic
algorithm that uses five positioning rules, five sorting rules, and two
rotation rules to create individuals and populations. Two positioning
rules outperformed the others, obtaining promising results compared with
the literature.},
keywords = {Strips, Footwear industry, Search methods, Sociology, Packaging, Proposals, Statistics, Nesting, Continuous Rotation, Heuristics},
doi = {doi:10.1109/CEC45853.2021.9504813},
  notes = {Also known as \cite{9504813}},
)


@INPROCEEDINGS(Bolufe-Rohler:2021:CEC,
%xplor 25 Aug 2021
  author = {Antonio Bolufe-Rohler and Ye Yuan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Machine Learning for Determining the Transition Point in Hybrid
Metaheuristics},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1115--1122},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={High-level relay hybrids are among the most effective
metaheuristics in multiple domains. However, the relay aspect of
hybridization raises the problem of when to perform the transition from
one algorithm to the next. This problem becomes more relevant in
exploration-only exploitation-only hybrids, where each algorithm
specializes in a specific task and performs rather poorly in the other.
This paper presents a novel way of approaching the transition problem as
a classification problem. Different classifiers are trained and tested
on the MPS-CMAES hybrid; computational results are presented for the
CEC'13 benchmark. The performance of the machine learning based hybrid
confirms the effectiveness of the approach by achieving a significant
improvement over the original hybrid.},
keywords = {Machine learning algorithms, Sociology, Machine learning, Evolutionary computation, Benchmark testing, Classification algorithms, Relays, metaheuristics, machine learning, hybrid heuris-tics, Minimum Population Search, CMA-ES},
doi = {doi:10.1109/CEC45853.2021.9504905},
  notes = {Also known as \cite{9504905}},
)


@INPROCEEDINGS(Kaszuba:2021:CEC,
%xplor 25 Aug 2021
  author = {Piotr Kaszuba and Maciej Komosinski and Agnieszka Mensfelt},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Automated Development of Latent Representations for Optimization
of Sequences Using Autoencoders},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1123--1130},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, we propose an automated method for the
development of new representations of sequences. For this purpose, we
introduce a two-way mapping from variable length sequence
representations to a latent representation modelled as the bottleneck of
an LSTM (long short-term memory) autoencoder. Desirable properties of
such mappings include smooth fitness landscapes for optimization
problems and better evolvability. This work explores the capabilities of
such latent encodings in the context of optimization of 3D structures.
Various improvements are adopted that include manipulating the
autoencoder architecture and its training procedure. The results of
evolutionary algorithms that use different variants of automatically
developed encodings are compared.},
keywords = {Training, Correlation, Three-dimensional displays, Sociology, Evolutionary computation, Predictive models, Genetics, genetic representation, autoencoder, latent representation, fitness landscape, evolutionary design},
doi = {doi:10.1109/CEC45853.2021.9504910},
  notes = {Also known as \cite{9504910}},
)


@INPROCEEDINGS(Pavelski:2021:CEC,
%xplor 25 Aug 2021
  author = {Lucas Marcondes Pavelski and Marie-Eleonore Kessaci and Myriam Delgado},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Local Optima Network Sampling for Permutation Flowshop},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1131--1138},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This work presents an analysis of sampled Local Optima
Networks (LONs) on flowshop instances. LON metrics are addressed here to
scrutinize the performance of four sampling strategies, each one
resulting from the combination of a local search (iterative or best
improvement) and a perturbation operator (deconstruction with or without
local search on partial solutions). The results highlight the
superiority of iterative improvement and the advantages of exploiting
partial solutions, especially when they are combined to discover new
local optima regions on large problems. LON metrics are also useful for
predicting performance of optimization algorithms - in particular, this
work considers two Iterated Greedy variants when solving flowshop
instances. The predictive capacity of metrics calculated from sampled
LONs is evaluated here based on the R^2 and RMSE indicators. Results
show that the best sampling strategies provide LON metrics capable of
predicting 8percent and 7percent of variance for small and large flowshop
instances, respectively.},
keywords = {Measurement, Perturbation methods, Computational modeling, Evolutionary computation, Predictive models, Markov processes, Search problems, Flowshop, Fitness Landscape, Local Optima Networks, LON Sampling Strategies},
doi = {doi:10.1109/CEC45853.2021.9504690},
  notes = {Also known as \cite{9504690}},
)


@INPROCEEDINGS(Skvorc:2021:CEC,
%xplor 25 Aug 2021
  author = {Urban Skvorc and Tome Eftimov and Peter Korosec},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={The Effect of Sampling Methods on the Invariance to Function
Transformations When Using Exploratory Landscape Analysis},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1139--1146},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Exploratory Landscape Analysis is a methodology for
transforming the samples of an optimization problem into numerical
descriptors called landscape features. Since Exploratory Landscape
Analysis is sample based, recent studies have shown that the choice of
the method used to collect the problem samples can have an effect on the
calculation of landscape features.In our recent work, we have shown that
certain landscape features are not invariant to even simple function
transformations such as shifting or scaling. However, the analysis in
our previous work was conducted using only Latin Hypercube Sampling.
Since we are now aware that the choice of sampling method can affect the
calculation of landscape features, this paper expands upon our earlier
work by using a variety of different sampling methods. We show that
different sampling methods do indeed have an effect on the invariance of
landscape features, and present a list of landscape features that are
invariant under all of our chosen sampling methods.},
keywords = {Taxonomy, Transforms, Evolutionary computation, Benchmark testing, Sampling methods, Prediction algorithms, Hypercubes, Exploratory Landscape Analysis, Numerical Optimization, Sampling},
doi = {doi:10.1109/CEC45853.2021.9504739},
  notes = {Also known as \cite{9504739}},
)


@INPROCEEDINGS(Alissa:2021:CEC,
%xplor 25 Aug 2021
  author = {Mohamad Alissa and Kevin Sim and Emma Hart},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Neural Approach to Generation of Constructive Heuristics},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1147--1154},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Both algorithm-selection methods and hyper-heuristic methods
rely on a pool of complementary heuristics. Improving the pool with new
heuristics can improve performance, however, designing new heuristics
can be challenging. Methods such as genetic programming have proved
successful in automating this process in the past. Typically, these make
use of problem state-information and existing heuristics as components.
Here we propose a novel neural approach for generating constructive
heuristics, in which a neural network acts as a heuristic by generating
decisions. We evaluate two architectures, an Encoder-Decoder LSTM and a
Feed-Forward Neural Network. Both are trained using the decisions output
from existing heuristics on a large set of instances. We consider
streaming instances of bin-packing problems in a continual stream that
must be packed immediately in strict order and using a limited number of
resources. We show that the new heuristics generated are capable of
solving a subset of instances better than the well-known heuristics
forming the original pool, and hence the overall value of the pool is
improved w.r.t. both Falkenauer's performance metric and the number of
bins used.},
keywords = {Measurement, Navigation, Heuristic algorithms, Neural networks, Genetic programming, Evolutionary computation, Dynamic scheduling, Automatic Heuristics Generation, Hyper-Heuristics, Encoder-Decoder LSTM, Streaming Bin-packing},
doi = {doi:10.1109/CEC45853.2021.9504989},
  notes = {Also known as \cite{9504989}},
)


@INPROCEEDINGS(Cruz-Duarte:2021:CEC,
%xplor 25 Aug 2021
  author = {Jorge M. Cruz-Duarte and Ivan Amaya and Jose Carlos Ortiz-Bayliss and Nelishia Pillay},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Automated Design of Unfolded Metaheuristics and the Effect of
Population Size},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1155--1162},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Metaheuristics are a fairly standard approach for solving
optimisation problems due to their success, flexibility, and simplicity.
However, there is a plethora of metaheuristics available, with different
performance levels for various problems. This work proposes a
methodology for designing heuristic-based procedures to solve continuous
optimisation problems and study how the population size affects its
performance. The technique comprises the well-known Simulated Annealing
algorithm as a hyper-heuristic, and a heuristic sequence taken from
unfolding the conventional scheme of population-based metaheuristics
reported in the literature. Our results show that the proposed approach
is a reliable alternative for tackling optimisation problems. We find
exciting insights, according to our data, about this primary
implementation when varying the population size in different challenging
problems.},
keywords = {Heuristic algorithms, Sociology, Simulated annealing, Thermal management, Search problems, Reliability, Proposals, Hyper-heuristic, Dimensionality, Search Operators, Continuous Optimisation, Evolutionary Computation},
doi = {doi:10.1109/CEC45853.2021.9504879},
  notes = {Also known as \cite{9504879}},
)


@INPROCEEDINGS(Derbel:2021:CEC,
%xplor 25 Aug 2021
  author = {Bilel Derbel and Geoffrey Pruvost and Byung-Woo Hong},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Enhancing Moea/d with Escape Mechanisms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1163--1170},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, we investigate the design of escape mechanisms
within the state-of-the-art decomposition-based evolutionary
multi-objective Moea/d framework. We propose to track the number of
improvements made with respect to the single-objective sub-problems
defined by decomposition. This allows us to compute an estimated
sub-problem improvement probability which serves as an activation signal
for some solution perturbation mechanism to occur. We report the
benefits of such an approach by conducting a comprehensive experimental
analysis on a broad range of combinatorial bi-objective bit-string
landscapes with variable dimensions and ruggedness. Our empirical
findings provide evidence on the effectiveness of the proposed escape
mechanism and its ability in providing substantial improvement over
conventional Moea/d. Besides, we provide a detailed analysis of
parameters impact and anytime behavior in order to better highlight the
strength of the proposed techniques as a function of available budget
and problem characteristics.},
keywords = {Perturbation methods, Sociology, Evolutionary computation, Statistics, Optimization, Multi-objective optimization, decomposition},
doi = {doi:10.1109/CEC45853.2021.9504957},
  notes = {Also known as \cite{9504957}},
)


@INPROCEEDINGS(da-Silva:2021:CEC,
%xplor 25 Aug 2021
  author = {Gabriela Cavalcante {da Silva} and Elizabeth F. Wanner and Leonardo C. T. Bezerra and Thomas Stutzle},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Revisiting Pareto-Optimal Multi- and Many-Objective Reference
Fronts for Continuous Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1171--1178},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The performance assessment of multi-objective heuristic
algorithms is one of the most significant contributions from the
evolutionary optimization algorithms community. By contrast, performance
assessment in the context of many-objective optimization is still a
challenging, open research field. Recent advances have demonstrated
disagreements between Pareto-compliant performance metrics, and
indicated that reference fronts produced by benchmark generators of
Pareto-optimal fronts could be further improved. In this work, we
investigate these reference fronts with the help of multi-dimensional
visualization techniques and Pareto-monotonic archivers. Interestingly,
reference fronts produced by benchmark generators for DTLZ and WFG
continuous optimization problems show significant issues, even when only
three objectives are considered. Furthermore, given that input solution
sets for five-objective problems are not high-quality, archivers are
unable to output reasonable approximation fronts. We conclude that the
performance assessment of EMO algorithms needs to urgently address
reference front generation.},
keywords = {Measurement, Visualization, Heuristic algorithms, Instruments, Evolutionary computation, Benchmark testing, Approximation algorithms},
doi = {doi:10.1109/CEC45853.2021.9504952},
  notes = {Also known as \cite{9504952}},
)


@INPROCEEDINGS(Ming:2021:CEC,
%xplor 25 Aug 2021
  author = {Mengjun Ming and Rui Wang and Tao Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Investigating Constraint Relationship in Evolutionary
Many-Constraint Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1179--1184},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This study contributes to the treatment of numerous
constraints in evolutionary many-constraint optimization through
consideration of the relationships between pair-wise constraints. In a
conflicting relationship, the functional value of one constraint
increases as the value in another constraint decreases. In a harmonious
relationship, the improvement in one constraint is rewarded with
simultaneous improvement in the other constraint. In an independent
relationship, the adjustment to one constraint never affects the
adjustment to the other. Based on the different features, methods for
identifying constraint relationships are investigated, helping to
simplify many-constraint optimization problems (MCOPs). Additionally,
the transitivity of the relationships is further discussed, facilitating
the determination of the relationship in a new pair of constraints.},
keywords = {Evolutionary computation, Optimization, many-constraint optimization problems (MCOPs), constraint relationships, conflict, harmony, independence},
doi = {doi:10.1109/CEC45853.2021.9504839},
  notes = {Also known as \cite{9504839}},
)


@INPROCEEDINGS(Marquez-Vega:2021:CEC,
%xplor 25 Aug 2021
  author = {Luis A. Marquez-Vega and Jesus Guillermo Falcon-Cardona and Edgar Covantes Osuna},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Towards a More Balanced Reference Set Adaptation Method: First
Results},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1185--1192},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Reference sets are widely used by many multi-objective
evolutionary algorithms (MOEAs) to decompose the objective space, define
search directions, or calculate quality indicators (QIs) embedded into
the selection mechanisms. Well-known MOEAs adopt the generation of
uniformly distributed points on a unit simplex to construct such
reference sets. Although these mechanisms are useful for approximating
Pareto fronts with regular shapes, i.e., simplex-like shapes, they have
difficulties representing Pareto fronts with irregular geometries. To
overcome this drawback, many reference set adaptation methods have been
proposed so far. However, some adaptation methods present a degraded
performance on regular Pareto front shapes, while others promote a
balanced performance. Nevertheless, an extensive assessment has not been
made. In this paper, a MOEA based on the inverted generational distance
plus indicator, using an adaptive reference set, is used to study the
performance of well-known adaptation methods. Although an adaptation
method promotes a balanced performance on both regular and irregular
Pareto front shapes, results show some difficulties related to the
distribution of solutions in complex Pareto front shapes. The results of
this study allow detecting the main drawbacks of adaptation methods,
which can be addressed by using diversity-oriented selection mechanisms
in the generation of reference sets. Hence, these could impact the
generation of reference set-based MOEAs achieving good coverage,
convergence, and diversity regardless of the Pareto front shape.},
keywords = {Geometry, Shape, Evolutionary computation, Search problems, Convergence, Multi-objective evolutionary algorithms, reference set adaptation methods, irregular Pareto front shapes},
doi = {doi:10.1109/CEC45853.2021.9504769},
  notes = {Also known as \cite{9504769}},
)


@INPROCEEDINGS(Moshaiov:2021:CEC,
%xplor 25 Aug 2021
  author = {Amiram Moshaiov and Yosef Breslav and Eliran Farhi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Modal Multi-Objective Evolutionary Optimization for
Problems with Solutions of Variable-Length},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1193--1200},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper focuses on solving a special kind of multimodal
multi-objective optimization problems (MMOPs) in which solutions are of
variable length. First, problem definition and solution framework is
suggested to allow using standard multimodal multi-objective
evolutionary algorithms (MMEAs) to solve the considered type of
problems. Next, a real-life example of the considered type of problems
is suggested concerning optimal antennas' layout-allocation design for a
wireless communication network. Finally, a modification to NSGA-II is
suggested and employed to solve such layout problems. When compared with
other MMEAs, it is shown that the proposed algorithm provides not only
better solution diversity in the decision-space, but also solutions with
superior performance vectors. It is suggested here that this is
attributed to the type of archive that is used here.},
keywords = {Wireless communication, Layout, Evolutionary computation, Optimization, Standards, Antennas, multi-concept optimization, multi-modal multi-objective optimization, wireless communication network, location optimization, variable number of dimensions, variable length problem, MMEAs},
doi = {doi:10.1109/CEC45853.2021.9504705},
  notes = {Also known as \cite{9504705}},
)


@INPROCEEDINGS(Bernabe-Rodriguez:2021:CEC,
%xplor 25 Aug 2021
  author = {Amin V. {Bernabe Rodriguez} and Carlos A. {Coello Coello}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Empirical Study on the Use of the S-energy Performance
Indicator in Mating Restriction Schemes for Multi-Objective Optimizers},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1201--1208},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Mating restrictions have been used to improve the performance
of Multi-Objective Evolutionary Algorithms (MOEAs) by altering the way
in which parents are selected in the recombination step. Originally
proposed for single-objective optimization, mating restrictions have
been implemented in different MOEAs obtaining mixed results. However,
the role of mating restrictions in diversity management/maintenance and
in the proper balance between exploration and exploitation within MOEAs
has not been studied in sufficient detail in spite of its evident
importance. In this paper, we present an empirical study on the impact
of three new mating restrictions based on the s-energy performance
indicator. When obtaining each individual's contribution to the total
s-energy, we implicitly obtain vicinity information, since a high
contribution means that an individual is relatively close to at least
some other individual, i.e., it is in a crowded region. Conversely, an
individual with a low contribution is in a non-crowded region. Using
this information we explore different strategies aiming to improve the
diversity of the population during its execution, as well as exploiting
the least crowded regions of the objective space. One of the main
advantages of our proposal are both its simplicity and its ability to
scale up (in objective function space). We evaluate the impact of our
proposals by implementing them in NSGA-III and comparing the obtained
results with respect to those of the original algorithm. Our
experimental results show that the use of mating restrictions does
provide improvements in most of the test instances adopted for some of
our proposed strategies.},
keywords = {Sociology, Evolutionary computation, Linear programming, Proposals, Statistics, Optimization, Convergence, Multiobjective Optimization, Evolutionary Algorithms, Mating Restrictions},
doi = {doi:10.1109/CEC45853.2021.9504869},
  notes = {Also known as \cite{9504869}},
)


@INPROCEEDINGS(Oussalem:2021:CEC,
%xplor 25 Aug 2021
  author = {Zakaria Oussalem and Rochan Avlur Venkat and Jahnavi Malagavalli and Arya Kumar Bhattacharya},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Comparative Performances of Neural Networks of Variant
Architectures Trained with Backpropagation and Differential Evolution},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1209--1216},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Work on Neuro-evolution has of late tended to focus on the
design of alternative architectures and hyper-parameters for different
classes of Neural Networks to improve performance and computational
efficiency, using Evolutionary Algorithms (EA). Predominantly, the
underlying mechanism for learning weight parameters remains traditional
backpropagation (BP), now considered a paradigm of Deep Learning (DL).
Many important facets of DL, like hierarchical construction of features
across layers in image recognition, vanishing gradients, etc., are taken
for granted without recognizing that these may implicitly be induced by
the properties of BP itself. EAs that perform global optimization in
contrast to local gradient descent of BP, if used extensively for ANN
training, can potentially disrupt these assumed facets of DL - and
construct alternative and interesting perspectives. There is a
surprising lack of research activity towards one-to-one performance
comparison between EA and BP, keeping precisely the same architectures,
activation functions and datasets, for complex regression and
classification problems. This work partially fills this gap. It is
demonstrated that Differential Evolution enhanced with local search, can
generate mildly better accuracies on regression problems on noisy
industrial datasets compared to standard BP solutions, and comparable
accuracies on popular image classification datasets. Consequently, this
establishes the baseline for revisiting the above-mentioned assumed
facets of DL, and potentially engender new and interesting perspectives.},
keywords = {Backpropagation, Training, Image recognition, Neural networks, Computer architecture, Evolutionary computation, Search problems, Neural Network architectures, optimizing weight parameters, backpropagation, Differential Evolution, local crossover and mutation, regression and classification, neuro-evolution},
doi = {doi:10.1109/CEC45853.2021.9504984},
  notes = {Also known as \cite{9504984}},
)


@INPROCEEDINGS(Shi:2021:CEC,
%xplor 25 Aug 2021
  author = {Rui Shi and Jianping Luo and Qiqi Liu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Fast Evolutionary Neural Architecture Search Based on Bayesian
Surrogate Model},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1217--1224},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Neural Architecture Search (NAS) is studied to automatically
design the deep neural network structure, freeing people from heavy
network design tasks. Traditional NAS based on individual performance
evaluation needs to train many networks generated by the search, and
compare the performance of the networks according to their accuracy,
which is very time-consuming. In this study, we propose to use a
two-category comparator based random forest model as a surrogate to
estimate the accuracy of the networks. thereby reducing heavy network
training process and greatly saving search time. Instead of directly
predicting the accuracy of each network, we propose to compare the
relative performance between each two networks in our proposed
two-category comparator. Furthermore, we implement the modeling process
of the surrogate model in the sampling space of the original training
data, which further accelerates the search process of the network in the
NAS. Experimental results show that our proposed NAS framework can
greatly reduce the search time, while the accuracy of the obtained
network is comparable to that of other state-of-the art NAS algorithms.},
keywords = {Training, Performance evaluation, Neural networks, Training data, Computer architecture, Evolutionary computation, Data models, Bayesian Optimization, Surrogate-assisted Evolutionary Algorithm, Neural Architecture Search, Deep Neural Network},
doi = {doi:10.1109/CEC45853.2021.9504999},
  notes = {Also known as \cite{9504999}},
)


@INPROCEEDINGS(Parsa:2021:CEC,
%xplor 25 Aug 2021
  author = {Maryam Parsa and Shruti R. Kulkarni and Mark Coletti and Jeffrey Bassett and J. Parker Mitchell and Catherine D. Schuman},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Objective Hyperparameter Optimization for Spiking Neural
Network Neuroevolution},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1225--1232},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Neuroevolution has had significant success over recent years,
but there has been relatively little work applying neuroevolution
approaches to spiking neural networks (SNNs). SNNs are a type of neural
networks that include temporal processing component, are not easily
trained using other methods because of their lack of differentiable
activation functions, and can be deployed into energy-efficient
neuromorphic hardware. In this work, we investigate two evolutionary
approaches for training SNNs. We explore the impact of the
hyperparameters of the evolutionary approaches, including tournament
size, population size, and representation type, on the performance of
the algorithms. We present a multi-objective Bayesian-based
hyperparameter optimization approach to tune the hyperparameters to
produce the most accurate and smallest SNNs. We show that the
hyperparameters can significantly affect the performance of these
algorithms. We also perform sensitivity analysis and demonstrate that
every hyperparameter value has the potential to perform well, assuming
other hyperparameter values are set correctly.},
keywords = {Training, Sensitivity analysis, Neuromorphics, Neural networks, Sociology, Evolutionary computation, Hardware, spiking neural networks, neuromorphic computing, evolutionary algorithms},
doi = {doi:10.1109/CEC45853.2021.9504897},
  notes = {Also known as \cite{9504897}},
)


@INPROCEEDINGS(Mousavirad:2021:CEC,
%xplor 25 Aug 2021
  author = {Seyed Jalaleddin Mousavirad and Diego Oliva and Salvador Hinojosa and Gerald Schaefer},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Differential Evolution-based Neural Network Training
Incorporating a Centroid-based Strategy and Dynamic Opposition-based
Learning},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1233--1240},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Training multi-layer neural networks (MLNNs), a challenging
task, involves finding appropriate weights and biases. MLNN training is
important since the performance of MLNNs is mainly dependent on these
network parameters. However, conventional algorithms such as
gradient-based methods, while extensively used for MLNN training, suffer
from drawbacks such as a tendency to getting stuck in local optima.
Population-based metaheuristic algorithms can be used to overcome these
problems. In this paper, we propose a novel MLNN training algorithm,
CenDE-DOBL, that is based on differential evolution (DE), a
centroid-based strategy (Cen-S), and dynamic opposition-based learning
(DOBL). The Cen-S approach employs the centroid of the best individuals
as a member of population, while other members are updated using
standard crossover and mutation operators. This improves exploitation
since the new member is obtained based on the best individuals, while
the employed DOBL strategy, which uses the opposite of an individual,
leads to enhanced exploration. Our extensive experiments compare
CenDE-DOBL to 26 conventional and population-based algorithms and
confirm it to provide excellent MLNN training performance.},
keywords = {Training, Heuristic algorithms, Sociology, Evolutionary computation, Nonhomogeneous media, Classification algorithms, Statistics, neural network training, optimisation, differential evolution, center-based strategy, dynamic opposition-based learning},
doi = {doi:10.1109/CEC45853.2021.9504801},
  notes = {Also known as \cite{9504801}},
)


@INPROCEEDINGS(Andersen:2021:CEC,
%xplor 25 Aug 2021
  author = {Hayden Andersen and Sean Stevenson and Tuan Ha and Xiaoying Gao and Bing Xue},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolving Neural Networks for Text Classification using Genetic
Algorithm-based Approaches},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1241--1248},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Convolutional Neural Networks (CNNs) have been well-known for
their promising performance in text classification and sentiment
analysis because they can preserve the 1D spatial orientation of a
document, where the sequence of words is essential. However, designing
the network architecture of CNNs is by no means an easy task, since it
requires domain knowledge from both the deep CNN and text classification
areas, which are often not available and can increase operating costs
for anyone wishing to implement this method. Furthermore, such domain
knowledge is often different in different text classification problems.
To resolve these issues, this paper proposes the use of Genetic
Algorithm to automatically search for the optimal network architecture
without requiring any intervention from experts. The proposed approach
is applied on the IMDB dataset, and the experimental results show that
it achieves competitive performance with the current state-of-the-art
and manually-designed approaches in terms of accuracy, and also it
requires only a few hours of training time.},
keywords = {Training, Text categorization, Neural networks, Sociology, Manuals, Network architecture, Convolutional neural networks, Genetic Algorithm, Convolutional Neural Network, Text Classification},
doi = {doi:10.1109/CEC45853.2021.9504920},
  notes = {Also known as \cite{9504920}},
)


@INPROCEEDINGS(Vieira:2021:CEC,
%xplor 25 Aug 2021
  author = {Carlos Vieira and Leslie Perez Caceres and Leonardo C. T. Bezerra},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evaluating Anytime Performance on {NAS-Bench-101}},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1249--1256},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Neural architecture search (NAS) is a field where
computational effort poses a significant challenge, requiring large
computing clusters and specialized hardware. Furthermore, the lack of
common experimental guidelines often compromises NAS comparison or
induces premature conclusions. In a recent work, NAS-Bench-101 was
proposed to help mitigate those factors, providing both a common
benchmark and experimental guidelines for its use. In this work, we
discuss the design choices in NAS-Bench-101 and propose improvements
that increase the potential of the benchmark. First, we bridge NAS and
the research on anytime performance, showing how a bi-objective
formulation of NAS can improve the insights provided by NAS-Bench-101.
Then, we discuss choices made in the design of the benchmark, namely (i)
the fixed-size encoding, (ii) the effects of the limited variability
available, (iii) the assessment of algorithms only from a TPU time
perspective, and; (iv) the number of repetitions proposed. We
demonstrate our contributions assessing the best-performing algorithms
originally benchmarked on NAS-Bench-101 and also irace, one of the
best-performing algorithm configurators from the literature. Results
indicate that (i) the anytime performance methodology enriches the
insights obtained from the assessment on the original NAS-Bench-101;
(ii) algorithm comparison is strongly affected by the design choices
discussed, and; (iii) the performance of SMAC in this benchmark is
significantly improved by our alternative setups.},
keywords = {Deep learning, Training, Computer architecture, Evolutionary computation, Benchmark testing, Encoding, Hardware},
doi = {doi:10.1109/CEC45853.2021.9504902},
  notes = {Also known as \cite{9504902}},
)


@INPROCEEDINGS(Mendes:2021:CEC,
%xplor 25 Aug 2021
  author = {Renan S. Mendes and Elizabeth F. Wanner and Flavio V. C. Martins and Kalyanmoy Deb},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Aggregation or Selection? Clustering Many Objectives for Vehicle
Routing Problem with Demand Responsive Transport},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1257--1264},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper discusses a dimensionality reduction procedure to
tackle a many-objective formulation of a Vehicle Routing Problem with a
Demand Responsive Transport (VRPDRT). The problem formulation presents
eight objective functions that aim to reduce the operating costs while
meeting passenger needs and providing a high-quality service. Two
different dimensionality reduction-based approaches, aggregation and
feature selection are employed to transform the many-objective
formulation into a bi-objective one. The reduction, applied during the
search evolution, follows a hierarchical clustering technique in which
the objective functions' similarity and conflict are explored. The
proposed approaches are compared with a classic version of MOEA/D that
solves the problem in its original formulation. Moreover, different
dimensionality reduction frequencies are tested to assess the impact on
the algorithms' performance. When comparing the outcomes in the original
objective space, the results show that the aggregation approach
outperforms the feature selection method, regardless of the
dimensionality reduction frequency. Furthermore, while there is no
statistical difference between the MOEA/D and the aggregation approach
and the MOEA/D outperforms the feature selection approaches.},
keywords = {Dimensionality reduction, Time-frequency analysis, Vehicle routing, Clustering algorithms, Transforms, Feature extraction, Linear programming, Many objective optimization, Dimensionality reduction, Aggregation, Feature selection},
doi = {doi:10.1109/CEC45853.2021.9504919},
  notes = {Also known as \cite{9504919}},
)


@INPROCEEDINGS(Smith:2021:CEC,
%xplor 25 Aug 2021
  author = {Marcellus Smith and Brandon Brown and Gerry Dozier and Michael King},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Mitigating Attacks on Fake News Detection Systems using
Genetic-Based Adversarial Training},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1265--1271},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The study of adversarial effects on AI systems is not a new
concept, but much of the research has been devoted to deep learning. In
this paper we explore the effects of adversarial examples on 4 machine
learning classifiers and measure the effectiveness of adversarial
training. Additionally, we present a novel method for selecting
adversarial training examples that lead to a more robust machine
learning system. Our results suggest that adversarial examples can
significantly hinder the classification performance and that adversarial
training is an effective defensive counter-measure.},
keywords = {Training, Deep learning, Evolutionary computation, Genetic algorithms, Adversarial Training, Adversarial Examples, Steady-State Genetic Algorithm, Adversarial Example Selection, GAT},
doi = {doi:10.1109/CEC45853.2021.9504723},
  notes = {Also known as \cite{9504723}},
)


@INPROCEEDINGS(Moreira:2021:CEC,
%xplor 25 Aug 2021
  author = {Leonardo Henrique Moreira and Celia {Ghedini Ralha}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evaluation of Decision-making Strategies for Robots in
Intralogistics Problems Using Multi-agent Planning},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1272--1279},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Robots are currently being used in many real-world problems,
especially intralogistics scenarios where manufacturing and warehousing
are continuously performed. Those robots can be affected by exogenous
events that cause failures during their execution, leading to a dynamic
environment condition. Thus, robots interact not only with the
environment but also with other agents, therefore they require a
coordination protocol to guarantee cooperation and recovery from
unexpected events. The intralogistics problem is a real-world
application to a multi-agent planning model where agents are seen as
autonomous mobile robots (AMR). Related work highlights the autonomy
level and investigates different strategies for applying a
decision-making process. However, there is a gap in the literature
regarding an analysis of the conditions that make a decentralized
decision-making process more suitable than a centralized control in a
dynamic environment. This study investigates how strongly connected AMRs
are, according to their nature, and the conditions and effects of the
actions they can execute. We contribute to the research area by
detailing how the analysis of features of the intralogistics scenario
can be useful to identify the most suitable decision-making strategy for
AMRs in such dynamic environments.},
keywords = {Measurement, Protocols, Robot kinematics, Decision making, Warehousing, Process control, Planning, real-world applications, dynamic environments, multi-agent planning},
doi = {doi:10.1109/CEC45853.2021.9504887},
  notes = {Also known as \cite{9504887}},
)


@INPROCEEDINGS(Shrestha:2021:CEC,
%xplor 25 Aug 2021
  author = {Kripash Shrestha and Rahul Dubey and Ashutosh Singandhupe and Sushil Louis and Hung La},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi Objective {UAV} Network Deployment for Dynamic Fire Coverage},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1280--1287},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Recent large wildfires and subsequent damage have increased
the importance of wildfire monitoring and tracking. However, human
monitoring on the ground or in the air may be too dangerous and we thus
investigate deploying Unmanned Aerial Vehicles (UAVs) to track
wildfires. Specifically, we attack the problem of distributed autonomous
control of UAVs using a set of potential fields to track wildfire
boundaries. A multiobjective evolutionary algorithm searches through the
space of potential field parameters to maximize fire coverage while
minimizing energy consumption. Fire spread is modelled by the well known
FARSITE fire model. Preliminary simulation results show that our
potential fields approach to UAV control leads to 10percent coverage of the
boundary by UAVs and 78.percent energy remaining on three testing scenarios.},
keywords = {Energy consumption, Simulation, Computational modeling, Fires, Evolutionary computation, Aerospace electronics, Unmanned aerial vehicles, Distributed control, multi objective optimization, fire tracking, UAV},
doi = {doi:10.1109/CEC45853.2021.9504947},
  notes = {Also known as \cite{9504947}},
)


@INPROCEEDINGS(Xie:2021:CEC,
%xplor 25 Aug 2021
  author = {Yue Xie and Aneta Neumann and Frank Neumann},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Heuristic Strategies for Solving Complex Interacting Large-Scale
Stockpile Blending Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1288--1295},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Stockpile blending problem is an important component of
mine production scheduling, where stockpiles are used to store and blend
raw material. The goal of blending material from stockpiles is to create
parcels of concentrate which contain optimal metal grades based on the
material available. The volume of material that each stockpile provides
to a given parcel is dependent on a set of mine schedule conditions and
customer demands. Therefore, the problem can be formulated as a
continuous optimization problem. In the real-world application, there
are several constraints required to guarantee parcels that meet the
demand of downstream customers. It is a challenge in solving the
stockpile blending problems since its scale can be very large. We
introduce two repaired operators for the problems to convert the
infeasible solutions into the solutions without violating the two tight
constraints. Besides, we introduce a multi-component fitness function
for solving the large-scale stockpile blending problem which can
maximize the volume of metal over the plan and maintain the balance
between stockpiles according to the usage of metal. Furthermore, we
investigate the well-known approach in this paper, which is used to
solve optimization problems over continuous space, namely the
differential evolution (DE) algorithm. The experimental results show
that the DE algorithm combined with two proposed duration repair methods
is significantly better in terms of the values of results than the
results on real-world instances for both one-month problems and
large-scale problems.},
keywords = {Schedules, Heuristic algorithms, Metals, Production, Evolutionary computation, Maintenance engineering, Raw materials, Heuristic strategies, optimization problem, evolutionary algorithm, mine production schedule problem},
doi = {doi:10.1109/CEC45853.2021.9504901},
  notes = {Also known as \cite{9504901}},
)


@INPROCEEDINGS(Francisquini:2021:CEC,
%xplor 25 Aug 2021
  author = {Rodrigo Francisquini and Tiago T. {da Silva} and Maria C. V. Nascimento},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Detecting Anomalies In Daily {COVID-19} Cases Data From Brazil
Capitals Using GSP Theory},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1296--1303},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The COVID-19 pandemic has created an urgency for studies to
understand the spread of the virus, in particular, to predict the number
of daily cases. This type of investigation depends heavily on the data
collected and made available manually. Therefore, data are susceptible
to human errors which can cause anomalies in the dataset. Understanding
and correcting anomalies in real-world application data is an important
task to ensure the reliability of the data analysis and prediction
tools. This paper presents a spectral anomaly detection and correction
strategy that uses concepts from the graph signal processing (GSP)
theory. The main advantage of the introduced strategy is to analyze the
variation in the daily number of cases with the proximity relation
between the investigated locations. Experiments were carried out with
real meteorological and mobility data for predicting the number of
COVID-19 cases by the classic prediction model known as autoregressive
integrated moving average exogenous (ARIMAX). Then, the anomaly
detection method was applied to determine the relationship between the
prediction errors and the anomalous variations identified by the tool.
The results show a strong relationship between the anomalous variations
and the errors made by the model and attest to the increase in the
accuracy of the prediction model after the normalization of the
anomalies.},
keywords = {COVID-19, Pandemics, Computational modeling, Evolutionary computation, Predictive models, Tools, Signal processing, Anomaly Detection, Graph Signal Processing, Forecasting},
doi = {doi:10.1109/CEC45853.2021.9504844},
  notes = {Also known as \cite{9504844}},
)


@INPROCEEDINGS(Varna:2021:CEC,
%xplor 25 Aug 2021
  author = {Fevzi Tugrul Varna and Phil Husbands},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Algorithm Assisted {HIDMS-PSO:} A New Hybrid Algorithm for
Global Optimisation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1304--1311},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, a new hybrid algorithm, GA-HIDMS-PSO, is
introduced by hybridising the state-of-the-art particle swarm
optimisation (PSO) variant, the heterogeneous improved dynamic
multi-swarm PSO (HIDMS-PSO) with a genetic algorithm (GA). The new
hybrid model exploits the heterogeneous features of HIDMS-PSO and the
evolutionary characteristics of the GA. In the GA-HIDMS-PSO
architecture, HIDMS-PSO acts as the primary search engine, and the GA is
employed as the secondary method to assist and slow down the loss of
diversity for selected proportions of homogeneous and heterogeneous
subpopulations of the HIDMS-PSO algorithm. Both methods run
consecutively. As the primary search method, HIDMS-PSO runs for longer
periods compared with the GA. The HIDMS-PSO pro-vides the initial
solutions for the GA from both homogeneous and heterogeneous
subpopulations and final solutions returned from the GA replace prior
solutions in the HIDMS-PSO which resumes the search process with
potentially more diverse particles to guide the swarm. The GA-HIDMS-PSO
algorithm's performance was tested on the 30 and 50 dimensional CEC'05
and CEC'17 test suites. The results were compared with 24 algorithms,
with 12 state-of-the-art PSO variants and 12 other metaheuristics.
GA-HIDMS-PSO outperformed all 24 comparison algorithms on both test
suites for both 30 and 50 dimensions.},
keywords = {Search methods, Heuristic algorithms, Sociology, Search engines, Noise measurement, Particle swarm optimization, Statistics, particle swarm optimisation, genetic algorithm, swarm intelligence, evolutionary algorithm, hybrid algorithm},
doi = {doi:10.1109/CEC45853.2021.9504852},
  notes = {Also known as \cite{9504852}},
)


@INPROCEEDINGS(Oyekan:2021:CEC,
%xplor 25 Aug 2021
  author = {John Oyekan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Objective Optimisation of Robotic Active Particle Swarms
for Continuous Repair of Large Scale High Value Structures},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1312--1318},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The manufacture and creation of large scale high value
structures has been done by humans for centuries. Examples include the
Egyptian pyramids, Bridges, Modern Skyscrapers to mention a few. These
structures are large but also provide a high value in terms of economy,
culture, display of prestige to mention a few. With advances in space
technology, we are bound to see these large scale high value structures
constructed in space. The vacuum of space present us with the challenges
of repairing these structures. This is due to the inhospitable and
dangerous environment of space. With increasing number of structures in
space, there is bound to be more debris created resulting in high impact
damages to these high value structures. Inspired by the biological blood
clotting process and biological active particles, in this work, we
propose the use of a swarm of live on artificial active particles for
the purposes of continuous and timely repair of these structures. We
tackle one of the challenges of artificial active particles research;
the ability to navigate in crowded and obstacle filled environments.
This challenge can be viewed from the perspective of a constrained
multi-objective optimisation problem in which a balance between
exploration of an environment and its exploitation needs to be achieved
while taking into consideration the various other constraints that apply
to an active particle. In this work, we show how artificial active
particles could avoid obstacles in their environment through the use of
an exploration mechanism and find damaged sites. Our results show that
as the ability to explore increases, the active particles are able to
navigate around obstacles and find a damaged site. However, there is a
limit to this.},
keywords = {Navigation, Space technology, Maintenance engineering, Robot sensing systems, Nanoscale devices, Biology, Sensors, Active Particles, Swarm Optimisation, Autonomous Repair, Multi-Objective, Brownian motion},
doi = {doi:10.1109/CEC45853.2021.9504749},
  notes = {Also known as \cite{9504749}},
)


@INPROCEEDINGS(Guo:2021:CEC,
%xplor 25 Aug 2021
  author = {Jinglei Guo and Miaomiao Shao and Shouyong Jiang and Xinyu Zhou},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Niche Based Multi-objective Particle Swarm Optimizer},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1319--1326},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The aim of multi-objective particle swarm optimizer (MOPSO) is
to find an accurate and well-distributed approximation of the true
Pareto Front (PF). The intrinsic character of PSO puts convergence
first, which can cause great loss of population diversity. How to
maintain the convergence and diversity simultaneously is an essential
issue for MOPSO. In this paper, we propose a niche based multi-objective
particle swarm optimizer (NMOPSO) to balance the convergence and
diversity. First, a niche based on the Euclidean distance is constructed
for each particle, then the leading particle is chosen out either from
the niche or from the whole swarm. After that, two position update
strategies are designed to update the position of each particle. The
position update strategies provide two guiding models for leaders, one
is utilizing the difference vector between the leader and the current
particle, the other is directly taking some components of leaders. Three
well-known test suites are employed to verify the performance of NMOPSO.
Compared with three popular MOPSOs, simulation results show that NMOPSO
performs better on most of test problems.},
keywords = {Simulation, Sociology, Euclidean distance, Benchmark testing, Particle swarm optimization, Statistics, Optimization, multi-objective optimization, particle swarm optimizer, niche, position update},
doi = {doi:10.1109/CEC45853.2021.9504907},
  notes = {Also known as \cite{9504907}},
)


@INPROCEEDINGS(Matsumoto:2021:CEC,
%xplor 25 Aug 2021
  author = {Kanta Matsumoto and Chihiro Ikuta},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{PSO} with Coupled Map Lattice and Worker Ant's Law},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1327--1332},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this study, we present two types of approaches for
developing of particle swarm optimization (PSO). Firstly, we introduce
coupled map lattice (CML) that is used a logistic map, to the moving
equation of PSO. We consider that the PSO obtains two parts which are
moving according to random vector and moving according similar vector by
the CML phenomena. Secondly, we introduce Worker ant's law to the moving
equation of PSO. The Worker ant's law is a variant of the Pareto's law.
The ratio of ants is 2:6:2 that include hard working ants, normal ants,
and not working ants, respectively. We give the different momentum term
for each swarm that is divided according to Worker ant's law. It is
considered that the search range can be expanded by the different
momentum term without the divergence of solution search. We confirm that
the CML and Worker ant's law improve the solution searching ability of
PSO by comparison with the general PSOs.},
keywords = {Lattices, Particle swarm optimization, Logistics, Particle swarm optimization, Coupled Map Lattice, Worker ant's law, Evolutionary algorithm, Optimization problem},
doi = {doi:10.1109/CEC45853.2021.9504693},
  notes = {Also known as \cite{9504693}},
)


@INPROCEEDINGS(Liu:2021:CEC,
%xplor 25 Aug 2021
  author = {Jun Liu and Tianyun Shi and Xiaoning Ma and Rui Xue and Min Liu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Optimal Train Speed Optimization under Several Safety Points by
the PSO Algorithm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1333--1340},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Considering the safety factors in the railway, this paper
mainly focuses on the optimal speed of high-speed train or locomotive
from the starting station to the ending station. Firstly, some concepts
in the railway safety, which include the risk source, hidden danger,
accident and safety warning point, are introduced to better understand
the railway safety factors. Secondly, the mathematical model including
objective function and the corresponding constraints is analyzed to
solve the optimal train speed problem. Thirdly, the PSO algorithm is
used to provide and design the optimal speed train from the starting
station to the ending station. Simulation results show that PSO
algorithm can handle with the optimal train speed problem and the
obtained optimal speed are provided to ensure the train safety in the
constant transportation time.},
keywords = {Simulation, Railway safety, Transportation, Evolutionary computation, Linear programming, Rail transportation, Safety, PSO algorithm, train, security risk, optimal speed},
doi = {doi:10.1109/CEC45853.2021.9504810},
  notes = {Also known as \cite{9504810}},
)


@INPROCEEDINGS(Wojcik:2021:CEC,
%xplor 25 Aug 2021
  author = {Krzysztof Wojcik and Tomasz Kulpa and Krzysztof Trojanowski},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Upper Bounds for Particle Location Variance Convergence Measures
in the Stochastic Model of Particle Swarm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1341--1348},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Properties of the Particle Swarm Optimization method depend on
the parameters configuring the behavior of single particles. In the
stochastic particle stability analysis, there were proposed recurrent
formulas of the expected value and the variance of particle location.
Recently, a particle variance convergence time measure was defined, and
for a selected set of configurations, a simple explicit formula for
particle location variance under the stagnation assumption was derived.
We propose an explicit formula for the upper bound of the particle
variance convergence time for the same selected set of configurations.
Furthermore, we propose an explicit formula for the upper bound for the
earlier defined measure of the particle location variance stasis time.
We visualize both upper bounds and verify them in simulations.},
keywords = {Weight measurement, Upper bound, Atmospheric measurements, Computational modeling, Stochastic processes, Particle measurements, Time measurement, PSO with inertia weight, order-2 stability conditions, Particle Location Variance Convergence Time, Particle Location Variance Stasis Time},
doi = {doi:10.1109/CEC45853.2021.9504884},
  notes = {Also known as \cite{9504884}},
)


@INPROCEEDINGS(Basilio:2021:CEC,
%xplor 25 Aug 2021
  author = {Samuel C.A. Basilio and Afonso C. C. Lemonge and Erica C.R. Carvalho},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Population Size Dynamic Reduction Criterion in {PSO} Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1349--1356},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The size of the population is extremely important when
executing a population-based algorithm. Its portioning impacts how much
the algorithm will have for exploration and exploitation. An excessively
large population can benefit exploration as opposed to exploitation.
With a population below ideal, exploration may be impaired, and the
algorithm may quickly converge to a local optimum. Unfortunately, the
choice of population size is often made empirically, where the user
experiences different values, several times, for different problems,
without any well-defined criteria, often drawing only on his experience.
This type of approach can under-use the algorithm, generating waste in
both computational cost and results. In this work, we improve and study
an approximation metamodel as a particle reduction criterion for
particle swarm algorithms. This metamodel considers that if two
particles are relatively close and with similar velocitys, they will
tend to the same solution, allowing one of these to be eliminated. Five
traditional benchmark problems in the literature in the field of
engineering applications were performed and the results analyzed.},
keywords = {Heuristic algorithms, Sociology, Benchmark testing, Approximation algorithms, Computational efficiency, Statistics, Particle swarm optimization, Particle Swarm Optimization, Metamodel-Based Optimization, Structural Optimization Problems},
doi = {doi:10.1109/CEC45853.2021.9504791},
  notes = {Also known as \cite{9504791}},
)


@INPROCEEDINGS(Yu:2021:CEC,
%xplor 25 Aug 2021
  author = {Jun Yu and Zitong Wang and Yan Pei},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Cooperative Chaotic Evolution},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1357--1364},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We introduce two new strategies into the conventional chaotic
evolution (CE) to facilitate the exchange of information between
individuals instead of searching independently, and propose a
cooperative CE (CoCE) with stronger performance. The first strategy
designs a new crossover operation to ensure that mutant individual's
exchange genes with other better individuals. Specifically, a mutant
individual does not cross with its parent individual, but with a
randomly selected individual whose fitness is not worse than that of the
parent individual. The second strategy introduces an additional
attraction generated by the randomly selected individual acting on the
mutant individual, which can allow the offspring individual to prefer
potential areas and speed up the search process. To evaluate the
performance of our proposed strategies, we configure CoCE and CE with
the exact same parameter settings and run them on 28 benchmark functions
from CEC 2013 test suites. Each benchmark function is run 30 times
independently on three different dimensions (i.e., 2-D, 10-D, and 30-D),
and we also apply the Friedman test and Holm's multiple comparison test
to check significant differences between CE and its three variants at
the termination condition. The experimental results confirmed that our
proposed CoCE has a faster convergence speed and higher convergence
accuracy, and the acceleration effect is more significant for high
dimensional optimization problems.},
keywords = {Sociology, Collaboration, Evolutionary computation, Benchmark testing, Real-time systems, Statistics, Adaptive control, Chaotic Evolution, Cooperative Chaotic Evolution, Evolutionary Computation, Optimization, Ergodicity},
doi = {doi:10.1109/CEC45853.2021.9504825},
  notes = {Also known as \cite{9504825}},
)


@INPROCEEDINGS(Delazeri:2021:CEC,
%xplor 25 Aug 2021
  author = {Gustavo Delazeri and Marcus Ritt},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Fast Heuristics for Traveling Salesman Problems with Multiple
Flying Sidekicks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1365--1371},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Speedy delivery of products is important for customers and the
optimization of the last mile has particular challenges. In recent years
drone-supported delivery has become increasingly feasible and therefore
a topic of intense research. In this paper we address a traveling
salesman problem where a truck has to serve customers and can use drones
as flying sidekicks for the delivery of parcels. The goal is to assign
customers to be either served by the truck or by some drone and schedule
all events in order to minimize the total completion time of whole
operation. We propose a heuristic for this problem, and introduce
several novel components, including a multi-armed bandit selection for
partitioning customers, an efficient matching-based heuristic for
assigning drones to customers, local searches to improve solutions, and
a fast greedy scheduler to estimate solution quality. In computational
experiments we show that the combined approach finds better solutions
than current methods, in less time.},
keywords = {Schedules, Heuristic algorithms, Evolutionary computation, Color, Traveling salesman problems, Routing, Search problems, traveling salesman, flying sidekicks, drones, heuristic algorithm, local search, matching heuristic},
doi = {doi:10.1109/CEC45853.2021.9504975},
  notes = {Also known as \cite{9504975}},
)


@INPROCEEDINGS(Adubi:2021:CEC,
%xplor 25 Aug 2021
  author = {Stephen A. Adubi and Olufunke O. Oladipupo and Oludayo O. Olugbara},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Configuring the Perturbation Operations of an Iterated Local
Search Algorithm for Cross-domain Search: A Probabilistic Learning
Approach},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1372--1379},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Hyper-heuristics are general-purpose heuristic search
methodologies for solving combinatorial optimization problems (COPs).
Research findings have revealed that hyper-heuristics still suffer
generalization issues as different strategies vary in performance from
an instance of a COP to another. In this paper, an approach based on
Iterated Local Search (ILS) is proposed to raise the level of generality
of hyper-heuristics on the problem domains of the HyFlex framework. The
proposed approach utilizes a probabilistic learning technique to
automatically configure the behavior of the ILS algorithm during the
perturbation stage of the optimization process. In the proposed method,
the mutation and ruin-recreate heuristics are treated as distinct
entities and the learning layer automatically determines the level of
utilization of these heuristic categories depending on the problem
domain being solved. The concept of double shaking is also presented
where a solution can be perturbed twice before the intensification
phase. Experimental results reveal the level of generality achieved by
the proposed method as it recorded a minimum formula one score of 30.0
on each tested problem domain. Direct comparison with a state-of-the-art
ILS-based approach also establishes the significance of the learning
layer added to the perturbation stage of the ILS metaheuristic. Finally,
analysis of the perturbation behavior of the hyper-heuristic leads to an
interesting conclusion concerning the type of low-level heuristics that
are highly beneficial and non-beneficial to a given problem domain.},
keywords = {Learning systems, Machine learning algorithms, Perturbation methods, Computational modeling, Machine learning, Evolutionary computation, Benchmark testing, iterated local search, selection hyper-heuristics, probabilistic learning, algorithm configuration},
doi = {doi:10.1109/CEC45853.2021.9504841},
  notes = {Also known as \cite{9504841}},
)


@INPROCEEDINGS(Alghamdi:2021:CEC,
%xplor 25 Aug 2021
  author = {Tami Alghamdi and Robert B. Heckendorn},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Evolutionary Computation Based Model for Testing Transfer
Learning Strategies},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1380--1389},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={To study how Transfer Learning (TL) works and what are
effective strategies for transfer learning, we propose to model the TL
process using Evolutionary Computation. EC provides a clear model for a
problem as searching through a set of potential solutions. We are able
to more easily control and measure problem difficulty, problem
similarity, and methods of information transfer and relate these to
success. As a proof of concept, we will use a static source problem and
three fixed target problems with simple known relationships (see Section
III). We compare the effectiveness of several ways to transfer knowledge
learned from solving one problem to solving the new problems in the
context of the relationship between the problems. This we hope will
demonstrate that using our EC model is a fruitful way to investigate TL.
The results show there is an improvement for using some sampled methods
representing the "learned knowledge" of the source problem S. Also, the
results show that the diversity of the transferred population has some
positive effect on finding the optimal solution depending on the
relationship between source and target problems.},
keywords = {Computational modeling, Sociology, Transfer learning, Evolutionary computation, Search problems, Statistics, Testing, Transfer Learning, evolutionary computation, genetic algorithms, diversity, generations, population, source problem, target problem},
doi = {doi:10.1109/CEC45853.2021.9504888},
  notes = {Also known as \cite{9504888}},
)


@INPROCEEDINGS(Terhar:2021:CEC,
%xplor 25 Aug 2021
  author = {Fynn Terhar and Christian Icking},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={The Sectional Art Gallery and an Evolutionary Algorithm for
Approaching Its Minimum Point Guard Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1390--1397},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We propose an extension of Art Galleries in Computational
Geometry towards a more real world application related definition, well
suited for Evolutionary Algorithms. We introduce two additional sections
inside the Art Gallery. One section defines locations where guards may
be placed in order to cover the Art Gallery. A second section shall be
observed by these guards. We show that our definition of a Sectional Art
Gallery also includes the Regular Art Gallery as a special case.
Furthermore, we present an Evolutionary Algorithm which can approach
Minimum Point Guard Problems in such a way, that high quality
approximations can be found in Sectional Art Galleries. Our algorithm is
influenced by Particle Swarm Optimizers and Particle Filters. Depending
on the application, relaxations of the problem can be handled by our
algorithm. This way, state of the art methods on optimally solving
Minimum Point Guard Problems in Regular Art Galleries can be
outperformed by multiple guards, still reaching more than 99.9percent of
coverage. By running our algorithm on public data sets, we show the
effectiveness of our approach in both Regular and Sectional Art
Galleries.},
keywords = {Computational geometry, Art, Filtering algorithms, Approximation algorithms, Particle filters, Particle swarm optimization, Art Gallery Problem, Minimum Point Guard Problem, Evolutionary Algorithm, Particle Filter, Particle Swarm Optimization, Metaheuristic},
doi = {doi:10.1109/CEC45853.2021.9504843},
  notes = {Also known as \cite{9504843}},
)


@INPROCEEDINGS(Varna:2021:CEC,
%xplor 25 Aug 2021
  author = {Fevzi Tugrul Varna and Phil Husbands},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{HIDMS-PSO} with Bio-inspired Fission-Fusion Behaviour and a Quorum
Decision Mechanism},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1398--1405},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this study, we propose a new variant of the HIDMS-PSO
algorithm with a bio-inspired fission-fusion behaviour and a quorum
decision mechanism (FFQ-HIDMS-PSO). In the new algorithm, units are
conceptualised as self-organising fission-fusion societies that
determine and adopt a suitable behaviour using unit-based quorum
decisions. The incorporation of the two bio-inspired mechanisms provide
"diversity aware" self-organising units that react to stagnation of
particles by adopting a suitable fission-fusion behaviour, leading to a
more efficient algorithm capable of maintaining significantly better
population diversity throughout the search. The performance of the
proposed algorithm was verified with three distinct experiments
conducted using CEC'17 and CEC'05 test suites at 30 and 50 dimensions,
comparing against 12 state-of-the-art metaheuristics and 12
state-of-the-art PSO variants. The proposed algorithm showed superior
performance in these experiments by outperforming all 24 algorithms in
all three experiments at 30 and 50 dimensions. The empirical evidence
suggests that the proposed method also maintains significantly superior
population diversity in comparison to the original HIDMS-PSO.},
keywords = {Biological system modeling, Sociology, Cultural differences, Statistics, Particle swarm optimization, particle swarm optimisation, swarm intelligence, meta-heuristics},
doi = {doi:10.1109/CEC45853.2021.9504755},
  notes = {Also known as \cite{9504755}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {Maiyue Chen and Ying Tan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Exponentially Decaying Explosion in Fireworks Algorithm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1406--1413},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Fireworks algorithm (FWA) as an efficient and robust swarm
intelligence algorithm can successfully deal with complex multi-modal
problems. In this paper, a novel new explosion operator called
exponentially decaying explosion is proposed to enhance the local search
ability of fireworks algorithm based on the principle of utilizing more
information. The proposed method takes the idea of guided mutation a
step further and dismantled the explosion process into an exponentially
decaying series of guided explosion. The FWA variant with this explosion
operator is called exponentially decaying fireworks algorithm (EDFWA).
Theoretical analysis proved the superiority of EDFWA in terms of
information utilization ratio compared with GFWA. Experimental results
showed that EDFWA not only surpassed LoTFWA in low dimensional
situations, but also exhibited powerful searching capability on 1000
dimensional high dimensional problems compared with multiple
representative optimizers specially designed for large-scale problems.},
keywords = {Fireworks algorithm, Benchmark testing, Search problems, Explosions, Particle swarm optimization, Optimization, Fireworks Algorithms, Large-scale Optimization, Multi-modal Optimization, Swarm Intelligence, Evolutionary Algorithm, Information Utilization, Exponential Decay},
doi = {doi:10.1109/CEC45853.2021.9504974},
  notes = {Also known as \cite{9504974}},
)


@INPROCEEDINGS(Papetti:2021:CEC,
%xplor 25 Aug 2021
  author = {Daniele M. Papetti and Daniel A. Ashlock and Paolo Cazzaniga and Daniela Besozzi and Marco S. Nobile},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={If You Can't Beat It, Squash It: Simplify Global Optimization by
Evolving Dilation Functions},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1414--1422},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Optimization problems represent a class of pervasive and
complex tasks in Computer Science, aimed at identifying the global
optimum of a given objective function. Optimization problems are
typically noisy, multi-modal, non-convex, non-separable, and often
non-differentiable. Because of these features, they mandate the use of
sophisticated population-based meta-heuristics to effectively explore
the search space. Additionally, computational techniques based on the
manipulation of the optimization landscape, such as Dilation Functions
(DFs), can be effectively exploited to either "compress" or "dilate"
some target regions of the search space, in order to improve the
exploration and exploitation capabilities of any meta-heuristic. The
main limitation of DFs is that they must be tailored on the specific
optimization problem under investigation. In this work, we propose a
solution to this issue, based on the idea of evolving the DFs.
Specifically, we introduce a two-layered evolutionary framework, which
combines Evolutionary Computation and Swarm Intelligence to solve the
meta-problem of optimizing both the structure and the parameters of DFs.
We evolved optimal DFs on a variety of benchmark problems, showing that
this approach yields extremely simpler versions of the original
optimization problems.},
keywords = {Semantics, Stochastic processes, Simulated annealing, Benchmark testing, Space exploration, Object recognition, Particle swarm optimization, computational intelligence, global optimization, dilation functions, fitness landscape, evolutionary algorithms, swarm intelligence, fuzzy self-tuning particle swarm optimization},
doi = {doi:10.1109/CEC45853.2021.9504708},
  notes = {Also known as \cite{9504708}},
)


@INPROCEEDINGS(Tangherloni:2021:CEC,
%xplor 25 Aug 2021
  author = {Andrea Tangherloni and Simone G. Riva and Simone Spolaor and Daniela Besozzi and Marco S. Nobile and Paolo Cazzaniga},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={The Impact of Representation on the Optimization of Marker Panels
for Single-cell RNA Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1423--1430},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The increasing number of single-cell transcriptomic and
single-cell RNA sequencing studies are allowing for a deeper
understanding of the molecular processes underlying the normal
development of an organism as well as the onset of pathologies. These
studies continuously refine the functional roles of known cell
populations, and provide their characterization as soon as putatively
novel cell populations are detected. In order to isolate the cell
populations for further tailored analysis, succinct marker
panels--composed of a few cell surface proteins and clusters of
differentiation molecules--must be identified. The identification of
these marker panels is a challenging computational problem due to its
intrinsic combinatorial nature, which makes it an NP-hard problem.
Genetic Algorithms (GAs) have been successfully used in Bioinformatics
and other biomedical applications to tackle combinatorial problems. We
present here a GA-based approach to solve the problem of the
identification of succinct marker panels. Since the performance of a GA
is strictly related to the representation of the candidate solutions, we
propose and compare three alternative representations, able to
implicitly introduce different constraints on the search space. For each
representation, we perform a fine-tuning of the parameter settings to
calibrate the GA, and we show that different representations yield
different performance, where the most relaxed representations-- in which
the GA can also evolve the number of genes in the panel--turn out to be
the more effective, especially in the case of 0-knowledge problems. Our
results also show that the marker panels identified by GAs can
outperform manually curated solutions.},
keywords = {Proteins, Sequential analysis, Pathology, RNA, Sociology, Color, Statistics, Candidate Solution Representation, Genetic Algorithms, Single-Cell Transcriptomic Data, Computational Biology, Bioinformatics},
doi = {doi:10.1109/CEC45853.2021.9504808},
  notes = {Also known as \cite{9504808}},
)


@INPROCEEDINGS(Pitzer:2021:CEC,
%xplor 25 Aug 2021
  author = {Erik Pitzer and Michael Affenzeller},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Cheating Like The Neighbors: Logarithmic Complexity For Fitness
Evaluation In Genetic Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1431--1438},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={While neighborhood-based algorithms can use incremental
evaluation to obtain the fitness of a modified solution candidate,
genetic crossover makes changes that are too big to easily allow reusing
previous quality values. In this paper, we extend our previous work and
evaluate the possibility of extending persistent data structures to
carry residual fitness values that can be reused for later evaluation of
derived solution candidates when applied to Multidimensional 0-1
Knapsack Problems. We show potential speedups especially on very large
problem instances when compared to classic array-based implementations.},
keywords = {Evolutionary computation, Genetics, Encoding, Complexity theory, Arrays, Optimization, Genetic algorithms, persistent data structures, metaheuristics, algorithms, optimization, genetic algorithms, crossover},
doi = {doi:10.1109/CEC45853.2021.9504849},
  notes = {Also known as \cite{9504849}},
)


@INPROCEEDINGS(Sharma:2021:CEC,
%xplor 25 Aug 2021
  author = {Shreya Sharma and Julian Blank and Kalyanmoy Deb and Bijaya Ketan Panigrahi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Ensembled Crossover based Evolutionary Algorithm for Single and
Multi-objective Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1439--1446},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A unique way evolutionary algorithms (EAs) are different from
other search and optimization methods is their recombination operator.
For real-parameter problems, it takes two or more high-performing
population members and blends them to create one or more new solutions.
Many real-parameter recombination operators have been proposed in the
literature. Each operator involves at least a parameter that controls
the extent of exploration (diversity) of the generated offspring
population. It has been observed that different recombination operators
and specific parameters produce the best performance for different
problems. This fact imposes the user to use different operator and
parameter combinations for every new problem. While an automated
algorithm configuration method can be applied to find the best
combination, in this paper, we propose an Ensembled Crossover based
Evolutionary Algorithm (EnXEA), which considers a number of
recombination operators simultaneously. Their parameter values and
applies them with a probability updated adaptively in proportion to
their success in creating better offspring solutions. Results on
single-objective and multi-objective, constrained, and unconstrained
problems indicate that EnXEA's performance is close to the best
individual recombination operation for each problem. This alleviates the
use of expensive parameter tuning either adaptively or manually for
solving a new problem.},
keywords = {Sociology, Optimization methods, Evolutionary computation, Statistics, Tuning, Crossover, Recombination, Ensemble-based algorithm, Evolutionary algorithm},
doi = {doi:10.1109/CEC45853.2021.9504698},
  notes = {Also known as \cite{9504698}},
)


@INPROCEEDINGS(Parque:2021:CEC,
%xplor 25 Aug 2021
  author = {Victor Parque},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Tackling the Subset Sum Problem with Fixed Size using an Integer
Representation Scheme},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1447--1453},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Addressing the subset sum problem is relevant to study
resource management problems efficiently. In this paper, we study a new
scheme to sample solutions for the subset sum problem based on
swarm-based optimization algorithms with distinct forms of selection
pressure, the balance of exploration-exploitation, the multimodality
considerations, and a search space defined by numbers associated with
subsets of fixed size. Our experiments show that it is feasible to find
optimal subsets with few number of fitness evaluations, and that
Particle Swarm Optimization with Fitness Euclidean Ratio converges
faster to the global optima with zero variability over independent runs.
Since the search space is one-dimensional and friendly to
parallelization schemes, our work is potential to study further classes
of combinatorial problems using swarm-based optimization algorithms and
the representation based on numbers.},
keywords = {Search problems, Rendering (computer graphics), Resource management, Particle swarm optimization, Optimization, subset sum, optimization, particle swarm, differential evolution, integer representation, experiments},
doi = {doi:10.1109/CEC45853.2021.9504889},
  notes = {Also known as \cite{9504889}},
)


@INPROCEEDINGS(Junqueira:2021:CEC,
%xplor 25 Aug 2021
  author = {Paulo Pinheiro Junqueira and Ivan Reinaldo Meneghini and Frederico Gadelha Guimaraes},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Local Neighborhood-Based Adaptation of Weights in Multi-Objective
Evolutionary Algorithms Based on Decomposition},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1454--1461},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Multi-objective algorithms based on decomposition have become
popular for the reason that a uniform distribution of weight vectors may
result in a better distribution of solutions along the Pareto front.
However, for more complex Pareto fronts with irregular shapes, the
initial weight vectors may not be adequate. One alternative to overcome
this problem, is to adapt the weight vectors during the evolutionary
process. In this paper an adaptive version of Multi-Objective
Evolutionary Algorithm based on Decomposition (MOEA/D) is proposed to
change the weight vectors based on the concept of local neighborhoods.
The proposed method is called MOEA/D with local-neighborhood-based
adaptation (MOEA/D-LNA). The proposed method is compared against a
number of famous variants of MOEA/D in the literature. Initial
experimental results have shown promising effectiveness on problems with
irregular Pareto shapes.},
keywords = {Shape, Heuristic algorithms, Evolutionary computation, Standards, Multi-objective optimization, adaptive weight vector, evolutionary algorithm, decomposition},
doi = {doi:10.1109/CEC45853.2021.9504688},
  notes = {Also known as \cite{9504688}},
)


@INPROCEEDINGS(Takagi:2021:CEC,
%xplor 25 Aug 2021
  author = {Tomoaki Takagi and Keiki Takadama and Hiroyuki Sato},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Weight Vector Arrangement Using Virtual Objective Vectors in
Decomposition-based MOEA},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1462--1469},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This work proposes an arrangement method of weight vectors
using virtual objective vectors supplementing the Pareto front
estimation. In decomposition-based evolutionary multi-objective
optimization, weight vectors decompose the Pareto front. Appropriate
weight vector distribution depends on the Pareto front shape, which is
generally unknown before the search. Objective vectors of obtained
non-dominated solutions become a clue to estimate the Pareto front shape
and arrange an appropriate weight vector set. However, a sizeable
objective vector set is required for a high-quality Pareto front
estimation and weight vector arrangement. The proposed method generates
and utilizes a virtual objective vector set based on the objective
vectors of obtained non-dominated solutions and an extended weight
vector set for the Pareto front estimation. Experimental results using
benchmark problems with different Pareto front shapes show that the
virtual objective vectors generated from a limited number of actual
objective vectors contribute to improving the search performance of
decomposition-based evolutionary multi-objective optimization.},
keywords = {Shape, Estimation, Evolutionary computation, Benchmark testing, Search problems, Planning, Optimization, multi-objective optimization, evolutionary algorithm, weight vector arrangement},
doi = {doi:10.1109/CEC45853.2021.9504954},
  notes = {Also known as \cite{9504954}},
)


@INPROCEEDINGS(Maldonado:2021:CEC,
%xplor 25 Aug 2021
  author = {Hugo Monzon Maldonado and Saul Zapotecas-Martinez},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Dynamic Penalty Function within {MOEA/D} for Constrained
Multi-objective Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1470--1477},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={For more than a decade, the efficiency and effectiveness of
MOEA/D (Multi-Objective Evolutionary Algorithm based on Decomposition)
when solving complicated problems has been shown. Due to this, several
researchers have focused their investigations on MOEA/D's extensions
that can deal with CMOPs (Constrained Multi-objective Optimization
Problems). In this paper, we adhere to the MOEA/D framework, a simple
penalty function to deal with CMOPs. The penalty function is dynamically
adapted during the search. In this way, the interaction between feasible
and infeasible solutions is promoted. As a result, the proposed approach
(namely MOEA/D-DPF) extends MOEA/D to handle constraints. The proposed
approach performance is evaluated on the well-known CF test problems
taken from the CEC'2009 suite. Using convergence and feasibility
indicators, we compare the solutions produced by our algorithm against
those produced by state-of-the-art MOEAs. Results show that MOEA/D-DPF
is highly competitive and, in some cases, it performs better than the
MOEAs adopted in our comparative study.},
keywords = {Heuristic algorithms, Sociology, Evolutionary computation, Search problems, Linear programming, Approximation algorithms, Statistics},
doi = {doi:10.1109/CEC45853.2021.9504940},
  notes = {Also known as \cite{9504940}},
)


@INPROCEEDINGS(Salih:2021:CEC,
%xplor 25 Aug 2021
  author = {Adham Salih and Amiram Moshaiov},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Modified Decomposition Framework and Algorithm for Many-objective
Topology and Weight Evolution of Neural Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1478--1485},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper presents a modified decomposition framework to
support the Many-Objective Topology and Weight Evolution of Artificial
Neural Networks (MaO-TWEANNs). Next, an algorithm, which is termed
NEWS/D, is devised using the proposed framework. To validate its
optimization capabilities, a numerical study is carried out. The
performed numerical study includes demonstration problems ranging from
three to seven objectives for which the ideal points are known. Finally,
an additional numerical study is performed with respect to a possible
real-life application. The latter study suggests that evolving class
experts for multi-class classification problems could be enhanced using
NEWS/D in a non-intuitive approach.},
keywords = {Network topology, Artificial neural networks, Evolutionary computation, Computer architecture, Distance measurement, Topology, Classification algorithms, Artificial Neural Networks, Many-objective Optimization, Decomposition Approach, Neuro-evolution, Topology and Weight Evolution, Neural Architecture Search},
doi = {doi:10.1109/CEC45853.2021.9504953},
  notes = {Also known as \cite{9504953}},
)


@INPROCEEDINGS(Pang:2021:CEC,
%xplor 25 Aug 2021
  author = {Lie Meng Pang and Hisao Ishibuchi and Ke Shang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Using a Genetic Algorithm-based Hyper-heuristic to Tune {MOEA/D}
for a Set of Various Test Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1486--1494},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The multi-objective evolutionary algorithm based on
decomposition (MOEA/D) is one of the most popular algorithms in the
field of evolutionary multi-objective optimization (EMO). Even though
MOEA/D has been widely used in many studies, it is likely that the
performance of MOEA/D is not always optimized since the same MOEA/D
implementation is often used on various problems with different
characteristics. However, obtaining an appropriate implementation of
MOEA/D for a different problem is not always easy, since there exists a
wide variety of choices for the components and parameters in MOEA/D. In
this paper, we examine the use of a genetic algorithm-based
hyper-heuristic procedure to offline tune MOEA/D on a single test
problem, a set of similar test problems, and a set of various test
problems. A total of 26 benchmark test problems are used in our study.
Experimental results show that the MOEA/D tuned for a set of various
test problems does not always perform well. It is also shown that the
MOEA/D tuned for a single test problem and for a set of similar test
problems always has high performance. Our experimental results strongly
suggest the necessity of using a tuning procedure to obtain a different
MOEA/D implementation for a different type of problems.},
keywords = {Clustering methods, Clustering algorithms, Evolutionary computation, Benchmark testing, Feature extraction, Search problems, Tuning, Evolutionary multi-objective optimization (EMO), hyper-heuristics, genetic algorithm, parameter tuning, MOEA/D},
doi = {doi:10.1109/CEC45853.2021.9504748},
  notes = {Also known as \cite{9504748}},
)


@INPROCEEDINGS(Sato:2021:CEC,
%xplor 25 Aug 2021
  author = {Mikiko Sato and Yuji Sato and Mads Midtlyng and Minami Miyakawa},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Inconstant Update of Reference Point Value for Parallel and
Distributed MOEA/D},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1495--1502},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper proposes an update method to determine the
reference point and nadir point when speeding up multi-objective
evolutionary optimization algorithm based on decomposition (MOEA/D) with
parallel and distributed processing in a many-core environment. We have
already proposed a migration method that uses a virtual overlapping zone
between partitions. Compared with simply partitioning the objective
space or applying the standard island migration model, this method is
shown to be effective in improving the hypervolume (HV) value,
increasing diversity, and reducing regions where non-dominated solutions
do not exist between partitions (sparse areas). However, this proposed
method could not completely deal with the sparse areas' problem in the
early stage of solution searching. To resolve this problem, in addition
to the virtual overlap method, we propose here a method that switches
the update interval for determining the reference point and nadir point
by taking into account the progress of solution searching. To evaluate
our newly proposed method, we compare both convergence and diversity on
single-core MOEA/D and parallel MOEA/D using a two-objective constrained
knapsack problem, and show that our method is effective not only in
reducing sparse areas in the early stage but also in improving diversity
and increasing the HV value.},
keywords = {Distributed processing, Design methodology, Evolutionary computation, Search problems, Partitioning algorithms, Standards, Optimization, MOEA/D, parallel and distributed processing, reference point, multi-objective evolutionary algorithms},
doi = {doi:10.1109/CEC45853.2021.9504979},
  notes = {Also known as \cite{9504979}},
)


@INPROCEEDINGS(Gibson:2021:CEC,
%xplor 25 Aug 2021
  author = {Finley J. Gibson and Richard M. Everson and Jonathan E. Fieldsend},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Objective Bayesian Optimisation Using an Exploitative
Attainment Front Acquisition Function},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1503--1510},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Efficient methods for optimising expensive black-box problems
with multiple objectives can often themselves become prohibitively
expensive as the number of objectives is increased. We propose an infill
criterion based on the distance to the summary attainment front which
does not rely on the expensive hypervolume or expected improvement
computations, which are the principal causes of poor dimensional scaling
in current state-of-the-art approaches. By evaluating performance on the
well-known Walking Fish Group problem set, we show that our method
delivers similar performance to the current state-of-the-art. We further
show that methods based on surrogate mean predictions are more often
than not superior to the widely used expected improvement, suggesting
that the additional exploration produced by accounting for the
uncertainty in the surrogate's prediction of the optimisation landscape
is often unnecessary and does not aid convergence towards the Pareto
front.},
keywords = {Support vector machines, Legged locomotion, Uncertainty, Evolutionary computation, Predictive models, Linear programming, Fish, Expensive optimisation, Bayesian optimisation, infill criteria, acquisition functions},
doi = {doi:10.1109/CEC45853.2021.9504899},
  notes = {Also known as \cite{9504899}},
)


@INPROCEEDINGS(Peralta:2021:CEC,
%xplor 25 Aug 2021
  author = {Federico Peralta and Samuel Yanes and Daniel Gutierrez Reina and Sergio Toral},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Monitoring Water Resources through a Bayesian Optimization-based
Approach using Multiple Surface Vehicles: The Ypacarai Lake Case Study},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1511--1518},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Bayesian optimization is a popular sequential decision
strategy that can be used for environmental monitoring. In this work, we
propose an efficient multi-Autonomous Surface Vehicle system capable of
monitoring the Ypacarai Lake (San Bernardino, Paraguay) (60 km^2 ) using
the Bayesian optimization approach with a Voronoi Partition system. The
system manages to quickly approximate the real unknown distribution map
of a water quality parameter using Gaussian Processes as surrogate
models. Furthermore, to select new water quality measurement locations,
an acquisition function adapted to vehicle energy constraints is used.
Moreover, a Voronoi Partition system helps to distributing the workload
with all the available vehicles, so that robustness and scalability is
assured. For evaluation purposes, we use both the mean squared error and
computational efficiency. The results showed that our method manages to
efficiently monitor the Ypacarai Lake, and also provides confident
approximate models of water quality parameters. It has been observed
that, for every vehicle, the resulting surrogate model improves by 3percent.},
keywords = {Computational modeling, Scalability, Water quality, Gaussian processes, Lakes, Robustness, Bayes methods, Bayesian Optimization, Gaussian Processes, Data Acquisition, Environmental Monitoring, Multi-robot Informative Path Planning, Autonomous Vehicles, Voronoi Diagrams},
doi = {doi:10.1109/CEC45853.2021.9504704},
  notes = {Also known as \cite{9504704}},
)


@INPROCEEDINGS(Tsujino:2021:CEC,
%xplor 25 Aug 2021
  author = {Koki Tsujino and Tomohiro Harada and Ruck Thawonmas},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Adaptation of Search Generations in Extreme Learning Assisted
MOEA/D Based on Estimation Accuracy of Surrogate Model},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1519--1526},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In the last decade, multi-objective evolutionary algorithms
(MOEAs) have been utilized for many real-world applications. However, it
takes a great deal of computation time for the majority of real-world
problems to obtain the optimal solutions due to the expensive fitness
evaluation cost. In order to reduce the computation time for
optimization, surrogate-assisted MOEAs have been studied. Our previous
study analyzed ELMOEA/D, one of the surrogate-assisted MOEA combining
MOEA/D with an extreme learning machine (ELM), from the relation between
search performance and search generations. Our previous analysis
revealed that the search generations on the surrogate space must be
determined to make the accuracy of the surrogate model low. For this
fact, this paper proposes the automatic adjustment methods for the
search generations of ELMOEA/D. We conduct experiments with several
well-known multi-objective benchmark problems and compare the proposed
methods with the conventional ELMOEA/D with the fixed number of
generations. The experimental results reveal that the proposed methods
achieve a more stable search performance than ELMOEA/D with the fixed
number of generations regardless of the target problems.},
keywords = {Adaptation models, Estimation error, Extreme learning machines, Computational modeling, Neural networks, Evolutionary computation, Gaussian processes, Surrogate model, extreme learning machine, MOEA/D, multi-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504819},
  notes = {Also known as \cite{9504819}},
)


@INPROCEEDINGS(Cao:2021:CEC,
%xplor 25 Aug 2021
  author = {Chang Cao and Xiaofen Lu and Yachen Li and Junda Zhu and Ke Tang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={The Performance Effect of Model Accuracy on
Classification-Assisted Evolutionary Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1527--1536},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Optimization problems with costly function evaluation widely
exist in real-world applications. Surrogate models are commonly used in
the field of optimization to deal with such expensive optimization
problems. In surrogate model-assisted evolutionary algorithms (EAs),
surrogate models like regression models, ranking models or
classification models are built based on historical data and then used
to compare candidate solutions in place of real function evaluations.
Researchers have also proposed various methods to make better use of
surrogate models in the optimization process of EAs. However, there is
no comprehensive study about how much accuracy of the built model is
accurate enough to bring benefits to the optimization. Motivated by
this, this work proposes a method to study the performance effect of
model accuracy on surrogate model-assisted EAs. Specifically, the method
does not really build surrogate models but assumes different model
accuracies in individual selection. Two classification-assisted EAs,
classification-assisted differential evolution (CADE) and relationship
classification-based preselection strategy (RCPS) are analyzed in this
work. The experimental results on a set of test functions show that a
weak learner with classification accuracy larger than 5percent is acceptable
ignoring the cost of model building. Another observation is that the
performances of CADE and RCPS increase monotonically and nonlinearly
with the classification accuracy.},
keywords = {Computational modeling, Buildings, Evolutionary computation, Data models, Optimization, Evolutionary algorithms, surrogate model, differential evolution, classification, preselection},
doi = {doi:10.1109/CEC45853.2021.9504809},
  notes = {Also known as \cite{9504809}},
)


@INPROCEEDINGS(Yao:2021:CEC,
%xplor 25 Aug 2021
  author = {Yiming Yao and Xudong Yang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Efficient Global Multi-Objective Aerodynamic Optimization Using
Combined Multi-Point Infilling Strategy and Surrogate Models},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1537--1542},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Surrogate-based method can dramatically reduce the number of
expensive function evaluations in real-world multi- objective
optimization problems (MOP). When the number of objectives is small,
using surrogate models combined with expected hypervolume improvement
(EHVI) infill sampling criterion (ISC) has been proved to be efficient
to provide a set of solutions with good diversity and good proximity to
the Pareto front (PF) in aerodynamic shape optimization. However,
traditional hypervolume-based infilling strategies use only one kind of
ISC to generate one or multiple sample points, the advantages of various
kinds of ISC cannot be comprehensively utilized and the parallelization
is not easy to implement. This paper proposes a combined multi-point
infilling strategy based on Kriging models and develops an efficient
global multi-objective constrained optimization method (EGMOCO) to solve
multi- objective aerodynamic shape optimization with complex
constraints. Multiple sample points are generated by using four ISC
considering hypervolume at each iteration and then evaluated in
parallel. Firstly, the performance of EGMOCO is compared with that of
single criterion EHVI strategy on six benchmarks within the same
computational budget to prove its effectiveness, and then EGMOCO is
implemented in an aerodynamic shape optimization problem with complex
constraints. The result shows that EGMOCO has good performance in
balancing local exploitation and global exploration with faster
convergence rate and high robustness, the whole PF can be fully explored
in limited evaluations and the constraint handling is effective
especially for real-world problems with complex and nonlinear
constraints, the comprehensive aerodynamic performance of the airfoil is
greatly improved. It can be confirmed that Kriging-based multi-objective
optimization method combined with multi-point infilling strategy
performs better than single infilling criterion EHVI, since different
sample infilling criteria can complement with each other, both local
exploitation and global exploration can be considered and well balanced.},
keywords = {Constraint handling, Shape, Computational modeling, Optimization methods, Evolutionary computation, Benchmark testing, Aerodynamics, Aerodynamic Shape Optimization, Multi-Objective Optimization, Kriging, Hypervolume, Parallelization, Infill Sampling Criteria},
doi = {doi:10.1109/CEC45853.2021.9504946},
  notes = {Also known as \cite{9504946}},
)


@INPROCEEDINGS(Kely-de-Melo:2021:CEC,
%xplor 25 Aug 2021
  author = {Maisa {Kely de Melo} and Rodrigo T. N. Cardoso and Tales {Argolo Jesus}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Genetic Algorithm For Investment Tracking With Stochastic Model
Predictive Control},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1543--1550},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper proposes a genetic algorithm (GA) for the
stochastic model predictive control (SMPC) approach to investment
tracking in the presence of transaction costs and cardinality
constraints. The methodology is based on minimizing a stochastic measure
of the tracking error predicted for the next trading date. A conditional
value at risk (CVaR) measure is proposed to determine the replicating
portfolio's optimal composition, quantifying the average loss over a
subset of worst- case realizations. The investment performance is tested
and analyzed in an out-of-sample simulation using data from the
Brazilian Stock Exchange. Results indicate that the proposed approach
can effectively track the benchmark and satisfy the constraints.},
keywords = {Stochastic processes, Pricing, Benchmark testing, Tools, Prediction algorithms, Loss measurement, Stock markets, model predictive control, portfolio optimization, tracking investment, cardinality, transaction costs},
doi = {doi:10.1109/CEC45853.2021.9504714},
  notes = {Also known as \cite{9504714}},
)


@INPROCEEDINGS(Alibrahim:2021:CEC,
%xplor 25 Aug 2021
  author = {Hussain Alibrahim and Simone A. Ludwig},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Hyperparameter Optimization: Comparing Genetic Algorithm against
Grid Search and Bayesian Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1551--1559},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The performance of machine learning algorithms are affected by
several factors, some of these factors are related to data quantity,
quality, or its features. Another element is the choice of an
appropriate algorithm to solve the problem and one major influence is
the parameter configuration based on the problem specification.
Parameters in machine learning can be classified in two types: (1) model
parameters that are internal, configurable, and its value can be
estimated from data such as weights of a deep neural network; and (2)
hyperparameters, which are external and its values can not be estimated
from data such as the learning rate for the training of a neural
network. Hyperparameter values may be specified by a practitioner or
using a heuristic, or parameter values obtained from other problems can
be used etc., however, the best values of these parameters are
identified when the algorithm has the highest accuracy, and these could
be achieved by tuning the parameters. The main goal of this paper is to
conduct a comparison study between different algorithms that are used in
the optimization process in order to find the best hyperparameter values
for the neural network. The algorithms applied are grid search
algorithm, bayesian algorithm, and genetic algorithm. Different
evaluation measures are used to conduct this comparison such as accuracy
and running time.},
keywords = {Training, Machine learning algorithms, Neural networks, Prediction algorithms, Search problems, Time measurement, Bayes methods, Hyperparmeter optimization, Grid Search, Bayesian, Genetic Algorithm},
doi = {doi:10.1109/CEC45853.2021.9504761},
  notes = {Also known as \cite{9504761}},
)


@INPROCEEDINGS(Soares-Silva:2021:CEC,
%xplor 25 Aug 2021
  author = {Julio Cezar {Soares Silva} and Adiel {Teixeira de Almeida Filho}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Simulated {IMO-DRSA} Approach for Cognitive Reduction in
Multiobjective Financial Portfolio Interactive Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1560--1568},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Obtaining the optimal Pareto front in multiobjective portfolio
problems can be impractical when considering real- world constraints,
such as portfolio cardinality. Thus, it is interesting to consider
multiobjective evolutionary algorithms to solve this type of problem in
a reasonable time, but without quality guarantees. IMO-DRSA can reduce
the search space using the preference information of the investor until
the single most preferable solution is found. The problem is that there
is no evidence on how to reduce the number of representative portfolios
to minimize decision-maker's (DM) cognitive effort during the
interaction, taking the satisfaction of preferences in future
distributions of portfolio components returns into account. The
objective of this work was to provide a way to support the reduction of
the number of representative examples presented for the investor, while
regarding out-of-sample preference satisfaction of the examples that
compose a data table. A simulation approach was proposed to analyze and
compare methods that select a small and robust sample of representative
solutions to compose data tables.},
keywords = {Analytical models, Sociology, Rough sets, Evolutionary computation, Search problems, Data models, Time factors, Portfolio selection, Interactive multiobjective optimization, Evolutionary algorithms, Dominance-based rough set approach},
doi = {doi:10.1109/CEC45853.2021.9504906},
  notes = {Also known as \cite{9504906}},
)


@INPROCEEDINGS(De-Amorim:2021:CEC,
%xplor 25 Aug 2021
  author = {Thiago Wanderley {De Amorim} and Julio Cezar Soares Silva and Adiel Teixeira {De Almeida Filho}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evaluation of Index Tracking Portfolios During the {COVID-19}
Pandemic},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1569--1576},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Index tracking consists of mimicking a benchmark performance
with a portfolio formed by a subset of assets contained in the index.
Due to the cardinality constraint, obtaining an optimal solution for
this problem can be impractical as the number of stocks in the index
grows. Then, meta-heuristics, such as genetic algorithms, can obtain
good solutions in a reason- able time, making it possible for the
investor to run different configurations of the problem before deciding
to rebalance or not his portfolio. Also, to evaluate an investment
strategy, it is important to perform backtests considering different
risk scenarios, especially in crisis scenarios, with a high volatile
market. This work aims to analyze the integration of hybrid and pure
genetic algorithms and two optimization models in a high volatility
market scenario, the Brazilian market index IBOVESPA during the COVID-19
pandemic. We observed that the hybrid algorithms returned competitive
solutions, tracking IBOVESPA even closer than the CPLEX solution on the
linear model for a non-rebalancing strategy. However, they were not
competitive in a rebalancing strategy, with solutions presenting a gap
of more than 10percent relative to the general-purpose solver solution.},
keywords = {COVID-19, Analytical models, Pandemics, Heuristic algorithms, Evolutionary computation, Indexes, Portfolios, Index tracking, COVID-19, genetic algorithm, backtesting},
doi = {doi:10.1109/CEC45853.2021.9504828},
  notes = {Also known as \cite{9504828}},
)


@INPROCEEDINGS(Vargas-Perez:2021:CEC,
%xplor 25 Aug 2021
  author = {Victor Vargas-Perez and Manuel Chica and Oscar Cordon},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{IPOP-CMA-ES} and the Influence of Different Deviation Measures for
Agent-Based Model Calibration},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1577--1584},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Calibration is a crucial task on building valid models before
exploiting their results. This process consists of adjusting the model
parameters in order to obtain the desired outputs. Automatic calibration
can be performed by using an optimization algorithm and a fitness
function, which involves a deviation measure to compare the time series
coming from the model. In this paper, we apply a memetic IPOP-CMA-ES for
the calibration of an agent-based model and we study the effect of
different deviation measures in this calibration problem. Classical
metrics calculate the mean point-to-point error, but we also propose
using an extension of dynamic time warping, which considers trend series
evolution. In order to determine if calibrating with an specific metric
leads to better solutions, we carry out an exhaustive experimentation by
including statistical tests, analysis on the values of the calibrated
parameters, and qualitative results. Our results show IPOP-CMA-ES
obtains better performance than a genetic algorithm. In addition, MAE,
MAPE and Soft-DTW are the metrics which report best results, although we
get a similar behavior for all of them.},
keywords = {Memetics, Computational modeling, Time series analysis, Evolutionary computation, Market research, Time measurement, Calibration, model calibration, agent-based modeling, time se- ries, deviation measures, evolutionary algorithms},
doi = {doi:10.1109/CEC45853.2021.9504694},
  notes = {Also known as \cite{9504694}},
)


@INPROCEEDINGS(Sun:2021:CEC,
%xplor 25 Aug 2021
  author = {Lijun Sun and Chao Lyu and Yuhui Shi and Chin-Teng Lin},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multiple-Preys Pursuit based on Biquadratic Assignment Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1585--1592},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The multiple-preys pursuit (MPP) is the adversarial game
between predators and preys. If the capture of a prey is defined as that
it cannot move anymore due to the surrounding of predators, there are
two kinds of task allocations. One is about assigning which prey to
which group of predators so that all preys can be captured. The other is
about assigning which capturing position to which predator to encircle
the prey simultaneously. In this paper, the MPP is modeled as a dynamic
optimization problem and each its time step is solved in two stages.
Firstly, the first kind of task allocation problem is modeled as the
biquadratic assignment problem (BiQAP) and a MPP fitness function is
proposed for the evaluation of such BiQAP task allocations. In this way,
the MPP is transformed to several single-prey pursuit (SPP) problems.
Secondly, for each SPP, we extend the coordinated SPP strategy CCPSO-R
(cooperative coevolutionary particle swarm optimization for robots) to
its parallel version as PCCPSO-R to enable the parallel implicit
capturing position allocating by parallel observation, decision making,
and moving of predators. Through experiments of the current BiQAP
solvers on the task allocation, we improve the best one of them in
statistic based on the domain knowledge. Moreover, the advantages of
PCCPSO-R in the capturing efficiency over CCPSO-R is testified in the
MPP experiments.},
keywords = {Heuristic algorithms, Decision making, Sociology, Fuzzy neural networks, Search problems, Resource management, Task analysis, multiple-preys pursuit, biquadratic assignment problem (BiQAP), cooperative coevolutionary algorithm},
doi = {doi:10.1109/CEC45853.2021.9504823},
  notes = {Also known as \cite{9504823}},
)


@INPROCEEDINGS(Saunders:2021:CEC,
%xplor 25 Aug 2021
  author = {Amanda Saunders and Daniel Ashlock and Julie Greensmith},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Necrotic Behavioral Control of Agent Behavior in the Iterated
Prisoner's Dilemma},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1593--1600},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={As the Covid-19 pandemic of 2020 illustrates, con-trolling the
behavior of social agents is a difficult problem. This study examines
the potential for an immune-inspired technique called necrosis to steer
the behavior of agent populations that are evolving to play the iterated
version of the game prisoner's dilemma. A key factor in this is
detection of behavioral types. The use of a previously developed
technique for fingerprinting the behavior of game playing agents, even
complex ones, permits the modelling of control strategies with necrotic
behavioral control (NBC). NBC consists of reducing the fitness of agents
engaging in an unacceptable behavior. The impact of applying necrosis to
a number of agent behaviors is investigated. The strategies
always-defect, always-cooperate, and tit-for-two-tats are used as the
foci for behavior control by zeroing out the fitness of agents whose
behavior is similar to those agents. Our experiments demonstrate that
NBC changes the distribution of prisoner's dilemma strategies that arise
both when the focal strategy is changed and when the similarity radius
used to zero out agent fitness is changed. Filtration focused on the
strategy tit-for-two-tats has the largest impact on the evolution of
prisoner's dilemma strategies while always cooperate is found to have
the least.},
keywords = {Pandemics, Filtration, Sociology, Game theory, Evolutionary computation, Behavioral science, Prisoner's dilemma, evolved agents, behavioral control, artificial immune system},
doi = {doi:10.1109/CEC45853.2021.9504866},
  notes = {Also known as \cite{9504866}},
)


@INPROCEEDINGS(Suresh:2021:CEC,
%xplor 25 Aug 2021
  author = {Anirudh Suresh and Jaturong Kongmanee and Kalyanmoy Deb and Vishnu Naresh Boddeti},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-objective Coevolution and Decision-making for Cooperative
and Competitive Environments},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1601--1608},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Co-evolutionary algorithms involve two co-evolving
populations, each having its own set of objectives and constraints, that
interact with each other during function evaluation. Co-evolutionary
algorithms are of great interest in cooperative and competing games and
search tasks in which multiple agents having different interests are in
play. Despite a number of single-objective co-evolutionary studies,
there has been limited interest in multi-objective co-evolutionary
algorithms. A recent study has revealed that in addition to the
challenges associated with the development of an efficient algorithm, a
proper understanding of the conflicting objectives within a single
population and their interaction among objectives of the second
population becomes extremely difficult to comprehend. In this paper, we
extend the previous proof-of-principle multi-objective co-evolutionary
(MOCoEv) study in three important directions. First, we enhance MOCoEv's
ability to handle mixed cooperating and conflicting scenarios among
different players. Second, we propose an iterative multi-criterion
decision-making (MCDM) approach to demonstrate how, in an arms-race type
scenario, the most appropriate solution can be selected from the
obtained Pareto-optimal solution set iteratively. Third, we extend the
previous MOCoEv algorithm with a many-objective evolutionary algorithm
(NSGA-III) to make them applicable to three or more objectives for each
player. These three developments reveal better insights about the
intricate issues related to multiple objectives and decision-making for
co-evolutionary optimization and take MOCoEv a step closer to solving
more complex multi-player problems.},
keywords = {Sociology, Decision making, Evolutionary computation, Games, Iterative methods, Statistics, Task analysis, Co-evolutionary algorithm, multi-objective optimization, multi-criterion decision making, cooperative and competitive co-evolution},
doi = {doi:10.1109/CEC45853.2021.9504868},
  notes = {Also known as \cite{9504868}},
)


@INPROCEEDINGS(Androulakakis:2021:CEC,
%xplor 25 Aug 2021
  author = {Pavlos Androulakakis and Zachariah Fuchs},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={The Effects of Initial Condition Selection on the Evolution of a
Turn Circle Intercept Feedback Controller},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1609--1618},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We explore the effects of four different initial condition
selection methods on the evolution of a feedback controller for the Turn
Circle Intercept problem. In this problem, an Attacker with constant
speed and bounded turn rate attempts to intercept a constantly turning
Target in minimum time. This type of problem arises frequently in air
combat and formation maintenance scenarios for winged aircraft. In
general, calculating a feedback controller's true fitness requires
evaluation from an infinite number of initial conditions. Since this is
not practical for nonlinear systems such as this, we instead approximate
the fitness by averaging the feedback controller's performance from a
subset of initial conditions. How these initial conditions are selected
and where they are located within the state space can have a large
impact on the performance of the algorithm and resulting evolved
solutions. In this research, we aim to quantify these differences by
comparing four different methods for selecting initial conditions: Grid
Sampling, Uniform Random Sampling, Randomized Step-Back Sampling, and
Hybrid Random Sampling.},
keywords = {Integrated circuits, Heuristic algorithms, Sociology, Evolutionary computation, Maintenance engineering, Approximation algorithms, Turning},
doi = {doi:10.1109/CEC45853.2021.9505011},
  notes = {Also known as \cite{9505011}},
)


@INPROCEEDINGS(Maina-Kilaas:2021:CEC,
%xplor 25 Aug 2021
  author = {Amani R. Maina-Kilaas and Cynthia Hom and Kevin Ginta and George D. Montanez},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={The Predator's Purpose: Intention Perception in Simulated Agent
Environments},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1619--1626},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We evaluate the benefits of intention perception, the ability
of an agent to perceive the intentions and plans of others, in improving
a software agent's survival likelihood in a simulated virtual
environment. To model intention perception, we set up a multi-agent
predator and prey model, where the prey agents search for food and the
predator agents seek to eat the prey. We then analyze the difference in
average survival rates between prey with intention perception--knowledge
of which predators are targeting them--and those without. We find that
intention perception provides significant survival advantages in almost
all cases tested, agreeing with other recent studies investigating
intention perception in adversarial situations and environmental danger
assessment.},
keywords = {Analytical models, Computational modeling, Software algorithms, Virtual environments, Evolutionary computation, Predator prey systems, Software agents, awareness, intention perception, attention perception, simulation, predator-prey model, intention trilogy},
doi = {doi:10.1109/CEC45853.2021.9504885},
  notes = {Also known as \cite{9504885}},
)


@INPROCEEDINGS(Abbassi:2021:CEC,
%xplor 25 Aug 2021
  author = {Malek Abbassi and Abir Chaabani and Lamjed Ben Said and Nabil Absi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Approximation-based Chemical Reaction Algorithm for
Combinatorial Multi-Objective Bi-level Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1627--1634},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Multi-objective Bi-Level Optimization Problem (MBLOP) is
defined as a mathematical program where one multi-objective optimization
task is constrained with another one. In this way, the evaluation of a
single upper level solution necessitates the evaluation of the whole
lower level problem. This fact brings new complexities to the bi-level
framework, added to the conflicting objectives and their evaluation
which need a large number of Function Evaluations (FEs). Despite the
number of works dedicated to solve bi-level optimization problems, the
number of methods applied to the multi-objective combinatorial case is
much reduced. Motivated by these observations, we propose in this paper
an approximation-based version of our recently proposed Bi-level
Multi-objective Chemical Reaction Optimization (BMCRO), which we called
BMCROII. The approximation technique is adopted here as a surrogate to
the lower level leading then to generate efficiently the lower level
optimality. Our choice is justified by two main arguments. First, BMCRO
applies a Quick Non-Dominated Sorting Algorithm (Q-NDSA) with
quasi-linear computational time complexity. Second, the number of FEs
savings gained by the approximation technique can hugely improve the
whole efficiency of the method. The proposed algorithm is applied to a
new multi-objective formulation of the well-known Bi-level Multi Depot
Vehicle Routing Problem (BMDVRP). The statistical analysis demonstrates
the outperformance of our algorithm compared to prominent baseline
algorithms available in literature. Indeed, a large number of savings
are detected which confirms the merits of our proposal for solving such
type of NP-hard problems.},
keywords = {Statistical analysis, Vehicle routing, Approximation algorithms, Iron, Proposals, Time complexity, Task analysis, Chemical reaction optimization, bi-level multi-objective optimization, approximation technique, combinatorial optimization, computational cost},
doi = {doi:10.1109/CEC45853.2021.9504711},
  notes = {Also known as \cite{9504711}},
)


@INPROCEEDINGS(Wang:2021:CEC,
%xplor 25 Aug 2021
  author = {Bing Wang and Hemant Kumar Singh and Tapabrata Ray},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Comparing Expected Improvement and Kriging Believer for Expensive
Bilevel Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1635--1642},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Bilevel optimization refers to a specialized class of problems
where one optimization task is nested as a constraint within another.
Such problems emerge in a range of real-world scenarios involving
hierarchical decision-making, and significant literature exists on
classical and evolutionary approaches to solve them. However,
computationally expensive bilevel optimization problems remain
relatively less explored. Since each evaluation incurs a significant
computational cost, one can only perform a limited number of function
evaluations during the course of search. Surrogate-assisted strategies
provide a promising way forward to deal with such classes of problems.
Of particular interest to this study are the steady-state strategies
which carefully pre-select a promising solution for true evaluation
based on a surrogate model. The main aim of this paper is to compare two
widely adopted steady-state infill strategies -Kriging believer (KB) and
expected improvement (EI) - through systematic experiments within a
nested optimization framework. Our experiments on a set of benchmark
problems reveal some interesting and counter-intuitive observations. We
discuss some of the underlying reasons and believe that the findings
will inform further research on understanding and improving search
strategies for expensive bilevel optimization.},
keywords = {Systematics, Computational modeling, Decision making, Switches, Evolutionary computation, Search problems, Steady-state, Expensive bilevel optimization, surrogate modeling, evolutionary algorithm},
doi = {doi:10.1109/CEC45853.2021.9504815},
  notes = {Also known as \cite{9504815}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {Lei Chen and Hai-Lin Liu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Transfer Learning Based Evolutionary Algorithm for Bi-level
Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1643--1647},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Evolutionary algorithms for bi-level optimization suffer from
the low efficiency in dealing with the multiple lower level optimization
tasks required by the population based upper level search. In this
paper, we design a transfer learning based covariance matrix adaptation
evolution strategy (TL-CMA-ES) for bi-level optimization problems, where
the tasks of searching for multiple lower level optimal solutions are
conducted by a set of CMA-ES optimizers in a parallel manner.
Furthermore, a transfer learning strategy is introduced in the parallel
lower level CMA-ES search such that each CMA-ES optimizer can learn and
utilize useful features gained by its neighbours. Experimental
comparison and analysis are carried out to verify the effectiveness and
efficiency of the proposed method.},
keywords = {Simulation, Transfer learning, Sociology, Evolutionary computation, Search problems, Task analysis, Covariance matrices, Bi-level optimization, Transfer learning, Evolutionary algorithm, CMA-ES},
doi = {doi:10.1109/CEC45853.2021.9504861},
  notes = {Also known as \cite{9504861}},
)


@INPROCEEDINGS(Gottipati:2021:CEC,
%xplor 25 Aug 2021
  author = {Srivathsa Gottipati and Praveen Paruchuri},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Genetic Algorithm Approach to Compute Mixed Strategy Solutions
for General Stackelberg Games},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1648--1655},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Stackelberg games have found a role in a number of
applications including modeling market competition, identifying traffic
equilibrium, developing practical security applications and many others.
While a number of solution approaches have been developed for these
games in a variety of contexts that use mathematical optimization,
analytical analysis or heuristic based solutions, literature has been
quite sparse on the usage of Genetic Algorithm (GA) based techniques. In
this paper, we develop a GA based solution to compute high quality mixed
strategy solution for the leader to commit to in a General Stackelberg
Game (GSG) using a normal form game formulation. The leader faces
multiple types of followers with discrete utility functions where the
mixed strategy of the leader (but not the actual action taken in the
round) is known to the follower. Our experiments showcase that the GA
developed here performs well in terms of scalability and provides
reasonably good solution quality in terms of the average reward
obtained. Given that finding the optimal mixed strategy solution for
GSGs is NP-hard (and the optimal solution for leader lies in the mixed
strategy space), we believe that the solution approach presented here
can support further development of practical applications using GSGs.},
keywords = {Scalability, Games, Evolutionary computation, Benchmark testing, Security, Mathematical model, Faces, Genetic Algorithms, General Stackelberg Games, mixed strategy},
doi = {doi:10.1109/CEC45853.2021.9505000},
  notes = {Also known as \cite{9505000}},
)


@INPROCEEDINGS(Dao:2021:CEC,
%xplor 25 Aug 2021
  author = {Tran Cong Dao and Tran Huy Hung and Nguyen Thi Tam and Huynh Thi Thanh Binh},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multifactorial Evolutionary Algorithm For Minimum Energy Cost
Data Aggregation Tree In Wireless Sensor Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1656--1663},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In wireless sensor networks, the majority of data transmitted
by sensor nodes is repeated over and over, and performing processes on
them in many cases leads to increased power consumption and reduced
network lifetime. Data aggregation is one of the techniques in reducing
redundancy and improving energy efficiency; it also increases the
lifespan of wireless sensor networks. In this paper, we address the
issues of constructing the data aggregation tree that minimizes the
total energy cost of data transmissions for two types of networks:
without relay nodes and using relay nodes. Traditionally, evolutionary
algorithms focus on constructing data aggregation trees for either
without relay node networks or using relay nodes networks. Therefore, we
propose Potential individuals based Multi-factorial Evolutionary
Algorithm (P-MFEA) to solve both issues simultaneously. The proposed
scheme shows improved performance in terms of energy consumption.},
keywords = {Wireless sensor networks, Energy consumption, Power demand, Simulation, Redundancy, Evolutionary computation, Data aggregation, data aggregation tree, network lifetime, energy efficiency, multifactorial evolutionary algorithm},
doi = {doi:10.1109/CEC45853.2021.9504807},
  notes = {Also known as \cite{9504807}},
)


@INPROCEEDINGS(Thanh:2021:CEC,
%xplor 25 Aug 2021
  author = {Le Tien Thanh and Le {Van Cuong} and Ta Bao Thang and Huynh {Thi Thanh Binh}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Armed Bandits for Many-Task Evolutionary Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1664--1671},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Inspired by the ability of human multitasking, there is a
growing body of literature in the computational intelligence community
dedicated to solving multiple problems concurrently. One of the research
areas that have been receiving much attention in this topic is
evolutionary multitasking, which is able to solve multiple complicated
optimization problems together, yielding a better result than solving
them in isolation. However, researches on evolutionary multitasking
mostly focus on solving a small number of problems together. In an
attempt to improve evolutionary multitasking, we propose Many-Task
Multi-Armed Bandit Evolutionary Algorithm (Ma^2 BEA), including a new
structure for the evolutionary multitasking algorithm. It also adopts a
well-proven result of Multi-Armed Bandit (MAB) as the method to control
the adaptive knowledge exchange between different tasks. In particular,
the action of selecting which task to perform inter-task crossover is
learned and decided by the designed MAB agent. We verify that Ma^2 BEA
correctly learned the underlying relationship between tasks using the
simple 10-task benchmarks. Besides, Ma^2 BEA is compared with other
evolutionary many-tasking algorithms that have recently been proposed
using the Single-Objective Many-task benchmark from the WCCI 2020
Competition on Evolutionary Multi-task Optimization. Empirical results
show that Ma^2 BEA is competitive in terms of high solution quality and
reasonable execution time.},
keywords = {Evolutionary computation, Benchmark testing, Multitasking, Distance measurement, Proposals, Task analysis, Knowledge transfer, Evolutionary Algorithm, Many-task Optimization, Multi-task Optimization, Task Selection, Fitness-based Measurement, Multi-armed Bandits},
doi = {doi:10.1109/CEC45853.2021.9504691},
  notes = {Also known as \cite{9504691}},
)


@INPROCEEDINGS(Feng:2021:CEC,
%xplor 25 Aug 2021
  author = {Yinglan Feng and Liang Feng and Yaqing Hou and Kay Chen Tan and Sam Kwong},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{EMT-ReMO:} Evolutionary Multitasking for High-Dimensional
Multi-Objective Optimization via Random Embedding},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1672--1679},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Since multi-objective optimization (MOO) involves multiple
conflicting objectives, the high dimensionality of the solution space
has a much more severe impact on multi-objective problems than
single-objective optimization. Taking the advantage of random embedding,
some related works have been proposed to scale derivative-free MOO
methods to high-dimensional functions. However, with the premise of "low
effective dimensionality", a single randomly embedded subspace cannot
guarantee the effectiveness of obtained solutions. Taking this cue, we
propose an evolutionary multitasking paradigm for multi-objective
optimization via random embedding (EMT-ReMO) to enhance the efficiency
and effectiveness of current embedding-based methods in solving
high-dimensional optimization problems with low effective dimensions. In
EMT-ReMO, the target problem is firstly embedded into multiple
low-dimensional subspaces by using different random embeddings, aiming
to build up a multi-task environment for identifying the underlying
effective subspace. Then the implicit multi-objective evolutionary
multitasking is performed with seamless knowledge transfer to enhance
the optimization process. Experimental results obtained on six
high-dimensional MOO functions with or without low effective dimensions
have confirmed the effectiveness as well as the efficiency of the
proposed EMT-ReMO.},
keywords = {Evolutionary computation, Multitasking, Search problems, Problem-solving, Task analysis, Optimization, Knowledge transfer, High-Dimensional Optimization, Evolutionary Multitasking, Random Embedding, Knowledge Transfer},
doi = {doi:10.1109/CEC45853.2021.9504857},
  notes = {Also known as \cite{9504857}},
)


@INPROCEEDINGS(Cai:2021:CEC,
%xplor 25 Aug 2021
  author = {Ronghong Cai and Jianping Luo},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Task Learning for Multi-Objective Evolutionary Neural
Architecture Search},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1680--1687},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Neural architecture search (NAS) is an exciting new field in
automating machine learning. It can automatically search for the
architecture of neural networks. But the current NAS has extremely high
requirements for hardware equipment and time costs. In this work, we
propose a predictor based on Radial basis function neural network
(RBFNN) as a surrogate model of Bayesian optimization to predict the
performance of neural architecture. The existing work does not consider
the difficulty of directly searching for neural architectures that meet
the performance requirements of NAS in real-world applications.
Meanwhile, NAS needs to execute multiple times independently when facing
multiple similar tasks. Therefore, we further propose a multi-task
learning surrogate model with multiple RBFNNs. The model not only
functions as a predictor, but also learns knowledge of similar tasks
jointly. The performance of NAS is improved by processing multiple tasks
simultaneously. Also, the current NAS is committed to searching for very
high-performance networks and does not take into account that neural
architectures are limited by device memory during actual deployment. The
scale of architecture also needs to be considered. We use a
multi-objective optimization algorithm to simultaneously balance the
performance and the scale, and build a multi-objective evolutionary
search framework to find the Pareto optimal front. Once the NAS is
completed, decision-makers can choose the appropriate architecture for
deployment according to different performance requirements and hardware
conditions. Compared with existing NAS work, our proposed MT-ENAS
algorithm is able to find a neural architecture with competitive
performance and smaller scale in a shorter time.},
keywords = {Performance evaluation, Computational modeling, Training data, Computer architecture, Predictive models, Search problems, Prediction algorithms, neural architecture search, multi-task learning, surrogate model, multi-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504721},
  notes = {Also known as \cite{9504721}},
)


@INPROCEEDINGS(Cuong:2021:CEC,
%xplor 25 Aug 2021
  author = {Le Van Cuong and Tran {Thi Huong} and Huynh Thi {Thanh Binh}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multi-task Approach For Maximum Survival Ratio Problem In
Large-Scale Wireless Rechargeable Sensor Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1688--1695},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={With the breakthrough of electromagnetic power transfer
technology, wireless charging has emerged as a hopeful solution for the
energy provisioning problem in wireless sensor networks. One of the
prominent issues that affect the potential exploitation of this
technology is the charging scheduling problem. However, existing works
on this topic either focus mainly on using a single mobile charger for
the whole network or suffer from several common limitations such as
enforcing the chargers to visit all sensors or applying the rigid
full-charging scheme. Moreover, they rarely delve into maximizing the
survival nodes ratio, which impacts directly on the multi-hop
communication of the network. This paper addresses the charging
scheduling for multiple mobile chargers without the above limitations.
We first formulate a maximum survival ratio problem and prove its
NP-hardness. A charging scheme that exploits the advantages of the
multifactorial evolutionary algorithm is then proposed to optimize the
charging paths of all chargers simultaneously. We finally evaluate the
efficacy of the proposed algorithm through extensive simulations. The
experimental results demonstrate that our scheduling scheme provides
promising outcomes in terms of survival ratio and the traveling energy
of chargers.},
keywords = {Wireless communication, Wireless sensor networks, Schedules, Inductive charging, Evolutionary computation, Spread spectrum communication, Sensors, Wireless rechargeable sensor network, multiple mobile chargers, maximum survival ratio problem, multifactorial evolutionary algorithm, multi-task optimization},
doi = {doi:10.1109/CEC45853.2021.9504930},
  notes = {Also known as \cite{9504930}},
)


@INPROCEEDINGS(Xu:2021:CEC,
%xplor 25 Aug 2021
  author = {Mengfan Xu and Zexuan Zhu and Yutao Qi and Lei Wang and Xiaoliang Ma},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Adaptive Multi-objective Multifactorial Evolutionary Algorithm
Based on Mixture Gaussian Distribution},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1696--1703},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In recent decades, multi-objective multifactorial evolutionary
algorithm (MOMFEA) has become a very promising research direction. How
to achieve effective knowledge transfer between similar tasks is the key
issue to affect the performance of the algorithm. In this paper, an
adaptive MOMFEA (AMOMFEA) is proposed by exploiting the mixture Gaussian
distribution of the population distributions of related tasks to help
solve the target task. Wasserstein distance is used to measure the
inter-task relevance in that the weight coefficient in the mixture
distribution is proportional to the inter-task relevance. Experimental
results on benchmark problems validate the effectiveness and efficiency
of the proposed method in comparison with MOMFEA and NSGA-II.},
keywords = {Weight measurement, Computational modeling, Sociology, Evolutionary computation, Mixture models, Gaussian distribution, Task analysis, MFEA, MO-MFEA, mixture model, and knowledge transfer},
doi = {doi:10.1109/CEC45853.2021.9504928},
  notes = {Also known as \cite{9504928}},
)


@INPROCEEDINGS(Jin:2021:CEC,
%xplor 25 Aug 2021
  author = {Bo Jin},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-Objective A* Algorithm for the Multimodal Multi-Objective
Path Planning Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1704--1711},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper, we consider the multimodal multi-objective path
planning (MMOPP) optimization, which is the main topic of a special
session in IEEE CEC 2021. The MMOPP aims at finding all the Pareto
optimal paths from a start area to a goal area on a grid map, while
passing through several designated must-visit areas. We propose an
efficient approach based on the multi-objective A* algorithm to exactly
solve the MMOPP. Experiments are conducted on the official MMOPP test
suite to evaluate the performance of the proposed approach. We also show
the admissibility of the proposed approach so that the computational
results can be used as the standard answers to the MMOPP test suite.},
keywords = {Shape, Computational modeling, Evolutionary computation, Pareto optimization, Path planning, Computational efficiency, Standards, Combinatorial optimization, multimodal multi-objective path planning optimization, multi-objective A* algorithm, best-first search, informed search},
doi = {doi:10.1109/CEC45853.2021.9504943},
  notes = {Also known as \cite{9504943}},
)


@INPROCEEDINGS(Dong:2021:CEC,
%xplor 25 Aug 2021
  author = {Nanjiang Dong and Rui Wang and Tao Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A New Encoding Mechanism Embedded Evolutionary Algorithm for {UAV}
Route Planning},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1712--1718},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Evolutionary algorithms (EAs) are often applied to deal with
UAV route planning. The solution encoding is one of important factor in
designing effective EAs. In a traditional encoding mechanism, each
individual represents one route. The whole population then consists of a
number of routes. We argue that such an encoding is less effective in
route planning, and then proposed an alternative encoding mechanism in
which one individual represents only one navigation point. The whole
population then represents one route. This implicitly turns EAs into
single-point based search with high exploitation ability. To further
improve the exploration ability of algorithms using this new encoding, a
slightly modified differential evolution operator is applied. Combining
the modified DE operator and the new encoding mechanism, the performance
of the derived algorithm is significantly improved, obtaining much
better route planning results than DE with the traditional encoding
mechanism.},
keywords = {Three-dimensional displays, Navigation, Heuristic algorithms, Sociology, Collaboration, Evolutionary computation, Encoding, UAV, encoding, route planning, evolutionary algorithm, differential evolution},
doi = {doi:10.1109/CEC45853.2021.9504874},
  notes = {Also known as \cite{9504874}},
)


@INPROCEEDINGS(Zhao:2021:CEC,
%xplor 25 Aug 2021
  author = {Jiaqi Zhao and Zhijie Jia and Yong Zhou and Ruihao Zhang and Zeming Xie and Zikang Xu and Yuxin Li and Di Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Path Planning based on Multi-objective Topological Map},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1719--1726},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Recently, intelligent robot technology has developed rapidly.
As an important topic, path planning has attracted more and more
attention. However, the performance of multi-objective path planning is
limited by the scale of the problem, and the results of path planning
are generally not optimal. In this work, based on 12 problems defined in
multimodal multiobjective path planning optimization contest for CEC
2021 [1], we present a multi-objective path planning optimization
algorithm, which includes data preprocessing, multi-objective genetic
evolution path planning algorithm, and non-dominated sorting algorithm
with elite strategy. This algorithm flow can realize multi-objective
path planning and generate the optimal solution set (i.e. Pareto
solution set). We implement our algorithm flow to calculate the time
complexity and space complexity indicators. Experimental results on
problems indicate the proposed multi-objective path planning algorithm
can solve the optimal solution set in time. Due to the constraints of
the problem, the number of optimal solutions is different for different
problems. We show the validity of our method with experiments for path
planning. Finally, we visually present some experimental results
intuitively. The code is released at
https://github.com/zhangruihao/pathPlanning.},
keywords = {Data preprocessing, Evolutionary computation, Genetics, Path planning, Complexity theory, Time complexity, Optimization, Path planning, multi-objective optimization, topological map},
doi = {doi:10.1109/CEC45853.2021.9504789},
  notes = {Also known as \cite{9504789}},
)


@INPROCEEDINGS(Hu:2021:CEC,
%xplor 25 Aug 2021
  author = {Yi Hu and Jing Liang and Boyang Qu and Jie Wang and Yanli Wang and Panpan Wei},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolutionary Ensemble Learning Using Multimodal Multi-objective
Optimization Algorithm Based on Grid for Wind Speed Forecasting},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1727--1734},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Improving the accuracy of wind speed forecasting is essential
for the usage of wind energy. This paper proposes an evolutionary
ensemble learning (EEL) method, which consists of ensemble empirical
mode decomposition (EEMD), random vector functional link network (RVFL)
based ensemble learning, and grid-based multimodal multi-objective
evolutionary algorithm (MMOG). Based on MMOG, the proposed ensemble
learning model is improved in terms of accuracy. Several benchmark
forecast methods are compared with the proposed EEL model on 12 wind
speed forecasting datasets. The experiment results validate the
superiority of the proposed EEL model in wind speed forecasting.},
keywords = {Measurement, Renewable energy sources, Wind energy, Wind speed, Evolutionary computation, Solar energy, Predictive models, Wind Speed Forecasting, Evolutionary Ensemble Learning, Ensemble Empirical Mode Decomposition, Random Vector Functional Link Network, Multimodal Multi-objective Evolutionary Algorithm},
doi = {doi:10.1109/CEC45853.2021.9504754},
  notes = {Also known as \cite{9504754}},
)


@INPROCEEDINGS(Robinson:2021:CEC,
%xplor 25 Aug 2021
  author = {Demelza Robinson and Qi Chen and Bing Xue and Isabella Wagner and Michael Price and Paul Hume and Kai Chen and Justin Hodgkiss and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Particle Swarm Optimisation for Analysing Time-Dependent
Photoluminescence Data},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1735--1742},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Next-generation photovoltaic materials such as per-ovskites
and organic photovoltaics are promising candidates for cheap,
solution-processable solar cells, which have the environmental and
financial advantages compared to traditional silicon-based cells. To
realise commercial solar cells, the development of new materials to
improve the performance is needed. Time-resolved optical spectroscopy is
a powerful technique for photovoltaic materials that allows measurement
of reveal the energy-dependent dynamics of photoexcited species (charges
and excitons), which are responsible for the device performance.
Although time-resolved spectroscopy provides contains rich information,
the data analysis can be time-consuming and labour-intensive. Automated
data-processing is therefore an attractive proposition to facilitate
higher throughput. This paper describes a new application of
evolutionary computation technique - particle swarm optimisation (PSO) -
to parametrise time-resolved photoluminescence (PL) data. PSO is used to
convert time- and energy-resolved photoluminescence data into decay rate
distributions. From this, the excited state lifetimes can be elucidated
- a key parameter for the optimisation of photovoltaic performance. The
implementation of PSO in enhanced LumiML proved advantageous, yielding
considerable improvements over previous techniques LumiML by two orders
of magnitude.},
keywords = {Photovoltaic systems, Performance evaluation, Spectroscopy, Photovoltaic cells, Photoluminescence, Optical variables measurement, Optical materials, Time-resolved Optical Spectroscopy, Decay Rate Distributions, Particle Swarm Optimisation, Renewable Energy, Material Selection},
doi = {doi:10.1109/CEC45853.2021.9504908},
  notes = {Also known as \cite{9504908}},
)


@INPROCEEDINGS(Giovannitti:2021:CEC,
%xplor 25 Aug 2021
  author = {Eliana Giovannitti and Sayyidshahab Nabavi and Giovanni Squillero and Alberto Tonda},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Exploiting Artificial Swarms for the Virtual Measurement of
Backlash in Industrial Robots},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1743--1750},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The backlash is a lost motion in a mechanism created by gaps
between its parts. It causes vibrations that increase over time and
negatively affect accuracy and performance. The quickest and most
precise way to measure the backlash is to use specific sensors, that
have to be added to the standard equipment of the robot. However, this
solution is little used in practice because raises the manufacturing
costs. An alternative solution can be to exploit a virtual sensor, i.e.,
the information about phenomena that are not directly measured is
reconstructed by signals from sensors used for other measurements.This
work evaluates the use of bio-inspired swarm algorithms as the
processing core of a virtual sensor for the backlash of a robotic joint.
Swarm-based approaches, with their relatively modest occupation of
memory and low computational load, could be ideal candidates to solve
the problem. In this paper, we exploit four state-of-the-art swarm-based
optimization algorithms: the Dragonfly Algorithm, the Ant Lion
Optimizer, the Grasshopper Optimization Algorithm, and the Grey Wolf
Optimizer. The four candidate algorithms are compared on 20 different
datasets covering a range of backlash values that reflect an industrial
case scenario. Numerical results indicate that, unfortunately, none of
the algorithms considered provides satisfactory solutions for the
problem analyzed. Therefore, even if promising, these algorithms cannot
represent the final choice for the problem of interest.},
keywords = {Vibrations, Service robots, Focusing, Sensor phenomena and characterization, Robot sensing systems, Manufacturing, Biosensors, backlash, robotic manipulator, real-world, industry application, swarms-based approaches},
doi = {doi:10.1109/CEC45853.2021.9504962},
  notes = {Also known as \cite{9504962}},
)


@INPROCEEDINGS(Zheng:2021:CEC,
%xplor 25 Aug 2021
  author = {Zhicheng Zheng and Yanan Li and Xiaokang Lei and Xingguang Peng},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Optimizing Parameters of Self-Organizing Model for Swarm Robots
via Evolutionary Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1751--1758},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A swarm of robots is a typical self-organizing system.
Parameters of self-organizing model can regulate the interaction among
the individuals to achieve the desired collective motion. Thus, it makes
sense to optimize the parameters for better performance in
self-organization. The evolutionary algorithm (EA) is one promising
heuristic optimization algorithm. However, quantitative evaluation of a
self-organizing model is still an open scientific question. In this
paper, we proposed three metrics for collective motion from the
perspective of grouping, polarization and spatial distribution of group.
Then, we designed the fitness function with the proposed metrics to
bridge the gap between EA and parameter tuning of self-organizing model.
In the experimental study, we optimized two kinds of self-organizing
models using the differential evolution algorithm (DE) and validated the
effectiveness in simulation. Moreover, we validated the optimized
parameters on a swarm of physical robots. The experimental results show
that the proposed metrics are effective to quantitatively evaluate the
self-organizing model so that the DE performs well for the automatic
parameter tuning of controlling swarm robots.},
keywords = {Measurement, Graphical models, Computational modeling, Swarm robotics, Evolutionary computation, Trajectory, Robots, swarm intelligence, evolutionary algorithm, swarm robots},
doi = {doi:10.1109/CEC45853.2021.9504922},
  notes = {Also known as \cite{9504922}},
)


@INPROCEEDINGS(Cai:2021:CEC,
%xplor 25 Aug 2021
  author = {Qing Cai and Haojie Ang and Sameer Alam},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multiobjective Optimization Approach for Reducing Air Traffic
Collision Risk},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1759--1766},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Air transport contributes significantly to the globalization
and world economic. With the increasing demand for both passengers and
air cargo, future airspace may encounter unprecedented traffic pressure.
It is always the paramount commitment of air transport to ensure flying
safety. In the face of increasing traffic demand, it is pertinent to
investigate how to reduce en-route collision risk without compromising
the traffic demand. In this paper, we propose a multiobjective
optimization based method to reduce the technical vertical risk (TVR) by
controlling en-route air traffic speed. The suggested method
simultaneously optimizes two objectives. The first one aims to minimize
the TVR while the second tries to minimize the traffic delay. As the
modeled optimization problem is non-convex and the two objectives
conflict with each other, we therefore introduce two well-known
multiobjective evolutionary algorithms named NSGA-II and NSGA-III and
modify some of their operators to solve the proposed optimization
problem. Finally, we carry out experiments on sixteen real-world daily
traffic sample data that cover en-route flights within the Singapore
flight information region (FIR). Experiments demonstrate that by
optimizing the proposed problem using the introduced algorithms we
obtain a set of speed control suggestions each of which can reduce the
TVR for the Singapore FIR. This work will contribute both to strategical
and tactical air traffic management as the aviation players can make the
preferred choices based on the solutions yielded by the introduced
algorithms.},
keywords = {Finite impulse response filters, Atmospheric modeling, Velocity control, Merging, Globalization, Evolutionary computation, Safety},
doi = {doi:10.1109/CEC45853.2021.9504707},
  notes = {Also known as \cite{9504707}},
)


@INPROCEEDINGS(Saha:2021:CEC,
%xplor 25 Aug 2021
  author = {Sneha Saha and Leandro L. Minku and Xin Yao and Bernhard Senhoff and Stefan Menzel},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Exploiting Linear Interpolation of Variational Autoencoders for
Satisfying Preferences in Evolutionary Design Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1767--1776},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In the early design phase of automotive digital development,
one of the key challenges for the designer is to consider
multiple-criteria like aerodynamics and structural efficiency besides
aesthetic aspects for designing a car shape. In our research, we imagine
a cooperative design system in the automotive domain which provides
guidance to the designer for finding sets of design options or
well-performing designs for preferred search areas. In the present
paper, we focus on two perspectives for this multi-criteria
decision-making problem: First, a scenario without prior information
about design preferences, where the designer aims to explore the search
space for a diverse set of design alternatives. Second, a scenario where
the designer has a prior intuition on preferred solutions of interest.
For both scenarios, we assume that historic 3D car shape data exists,
which we can utilize to learn a compact low-dimensional design
representation based on a variational autoencoder (VAE). In contrast to
evolutionary multi-objective optimization approaches where starting
populations are randomly initialized, we propose to seed the population
more efficiently by exploiting the advantage of linear interpolation in
the latent space of the VAE. In our experiments, we demonstrate that the
multi-objective optimization converges faster and achieves a diverse set
of solutions. For the second scenario, when specifying design
preferences by weights, we improve on the weighted-sum method, which
simplifies the multi-objective problem and propose a strategy for
efficiently adapting the weights towards the preferred design solution.},
keywords = {Interpolation, Three-dimensional displays, Shape, Sociology, Search problems, Space exploration, Automobiles, multi-objective optimization, preference formulation, autoencoders},
doi = {doi:10.1109/CEC45853.2021.9504772},
  notes = {Also known as \cite{9504772}},
)


@INPROCEEDINGS(Jara-Ten-Kathen:2021:CEC,
%xplor 25 Aug 2021
  author = {Micaela {Jara Ten Kathen} and Isabel {Jurado Flores} and Daniel {Gutierrez Reina} and Alejandro {Tapia Cordoba}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Autonomous Monitoring System for Water Resources based on {PSO} and
Gaussian Process},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1777--1784},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Monitoring water resources represents a crucial task not just
for nature preservation worldwide but also for human survival.
Traditional methods, based on manual sampling routines, are inefficient
in the context of large bodies of water. For this reason the interest in
new monitoring approaches based on autonomous vehicles has increased
during the last few years. Among these, strategies based on fleets or
swarms of vehicles have been proven especially efficient. One of the
main components of these systems is the global mission planner, which is
responsible for determining the optimal movements. In this paper a novel
Particle Swarm Optimization (PSO), based on a surrogate model like a
Gaussian Process, is proposed to guide a fleet of Autonomous Surface
Vehicles (ASV). The proposed approach takes advantage of the uncertainty
provided by the Bayesian model to guide the movements of the swarm
towards unexplored areas of the search space. The proposed system has
been validated using a benchmark function that models a water quality
parameter, achieving better results than the original PSO algorithm.},
keywords = {Uncertainty, Computational modeling, Gaussian processes, Water quality, Manuals, Particle swarm optimization, Task analysis, Particle Swarm Optimization, Gaussian Process, Water Monitoring, Autonomous Surface Vehicles, Machine Learning},
doi = {doi:10.1109/CEC45853.2021.9504936},
  notes = {Also known as \cite{9504936}},
)


@INPROCEEDINGS(Przewozniczek:2021:CEC,
%xplor 25 Aug 2021
  author = {Michal W. Przewozniczek and Marcin M. Komarnicki},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Fitness Caching - From a Minor Mechanism to Major Consequences in
Modern Evolutionary Computation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1785--1791},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In the field of Evolutionary Computation, the main objective
is to find a high-quality solution to the considered problem. However,
the other important issue is to find a solution efficiently. Therefore,
evolutionary methods use various techniques to adjust to the problem and
reduce the amount of resources consumed during an optimization process.
One of the well-known and relatively simple techniques is storing the
already evaluated genotypes and their ratings. Whenever an evolutionary
method is to evaluate the fitness for a genotype that was already rated,
instead of re-evaluating it, a method may use the value that was stored
in the repository. Surprisingly, despite its simplicity, such a fitness
caching technique was shown to cause many important phenomena. When an
evolutionary method is stuck, fitness caching may cause such significant
fitness function evaluation (FFE) reduction that FFE will not be a
reliable resource consumption measure anymore. Moreover, fitness caching
may help in detecting the drop in the number of new solutions
investigated by a method. Thus, it may help in dynamic population-size
management. Such a consequence is far more sophisticated than a simple
FFE reduction. Therefore, in this paper, we investigate fitness caching
in more detail. We analyze its influence on chosen state-of-the-art
methods employed to solve well-known theoretical problems.},
keywords = {Sociology, Evolutionary computation, Size measurement, Dynamic scheduling, Reliability, Statistics, Standards, Fitness Function Evaluations Number Minimization, Genetic Algorithms, Computation Load Measurement},
doi = {doi:10.1109/CEC45853.2021.9504686},
  notes = {Also known as \cite{9504686}},
)


@INPROCEEDINGS(Wu:2021:CEC,
%xplor 25 Aug 2021
  author = {Chenwang Wu and Wenjian Luo and Nan Zhou and Peilan Xu and Tao Zhu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Algorithm with Multiple Fitness Functions for Generating
Adversarial Examples},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1792--1799},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Studies have shown that deep neural networks (DNNs) are
susceptible to adversarial attacks, which can cause misclassification.
The adversarial attack problem can be regarded as an optimization
problem, then the genetic algorithm (GA) that is problem-independent can
naturally be designed to solve the optimization problem to generate
effective adversarial examples. Considering the dimensionality curse in
the image processing field, traditional genetic algorithms in
high-dimensional problems often fall into local optima. Therefore, we
propose a GA with multiple fitness functions (MF-GA). Specifically, we
divide the evolution process into three stages, i.e., exploration stage,
exploitation stage, and stable stage. Besides, different fitness
functions are used for different stages, which could help the GA to jump
away from the local optimum.Experiments are conducted on three datasets,
and four classic algorithms as well as the basic GA are adopted for
comparisons. Experimental results demonstrate that MF-GA is an effective
black-box attack method. Furthermore, although MF-GA is a black-box
attack method, experimental results demonstrate the performance of MF-GA
under the black-box environments is competitive when comparing to four
classic algorithms under the white-box attack environments. This shows
that evolutionary algorithms have great potential in adversarial attacks.},
keywords = {Deep learning, Image processing, Computational modeling, Evolutionary computation, Jitter, Optimization, Genetic algorithms},
doi = {doi:10.1109/CEC45853.2021.9504790},
  notes = {Also known as \cite{9504790}},
)


@INPROCEEDINGS(Benecke:2021:CEC,
%xplor 25 Aug 2021
  author = {Tobias Benecke and Sanaz Mostaghim},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Tracking the Heritage of Genes in Evolutionary Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1800--1807},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper aims to introduce a methodology to trace the
influence of the initial population of an evolutionary algorithm to the
final population. The major challenge concerns tracking the heritage of
multiple parent operators. In this paper, we propose a new encoding for
tracking purposes. In addition, we propose modifications to the
corresponding metrics for measuring the impact of individuals. With this
approach, we provide several tools to not only track the influence of
the initial population on the results but also to study the effects of
different crossover and mutation operators. In our experiments, we
evaluate the differences between two selected crossover and mutation
operators and provide insight into the proposed approach.},
keywords = {Measurement, Heuristic algorithms, Sociology, Genomics, Evolutionary computation, Tools, Encoding, Traceability, Evolutionary Algorithm, Initial Population, Convergence Behaviour},
doi = {doi:10.1109/CEC45853.2021.9504916},
  notes = {Also known as \cite{9504916}},
)


@INPROCEEDINGS(Gao:2021:CEC,
%xplor 25 Aug 2021
  author = {Yuanliang Gao and Sheung-Hung Poon},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Genetic Algorithm for Scheduling Splittable Tasks with
Precedence Constraints},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1808--1816},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In many applications, the logic of the program can be
described using a task graph, where the data dependencies and execution
time of each task are described. These dependencies create precedence
constraints among tasks, which are requirements that some tasks must be
finished before some other tasks. Many efforts have been put into
scheduling parallelizable tasks that synchronically use multiple cores.
In some cases, the task can be chunked into smaller pieces and scheduled
independently, allowing further flexible schedules. However, it is
usually either assumed that such splitting has no overhead, or that
precedence constraints are not present, or that the user has to provide
the way of splitting. This paper addresses the problem where these
factors are considered together, that is scheduling splittable tasks
with precedence constraints, where splitting introduces an overhead and
the splitting of tasks are determined by the algorithm. The objective is
to minimize the makespan of the schedule. We first present a
mixed-integer quadratic program (MIQP) formulation of the problem. Then,
a genetic algorithm (GA) is devised and its performance is compared with
the MIQP solutions. We show that the genetic algorithm can produce
reasonably good schedules compared with MIQP output within a
significantly shorter time, and it has the potential to handle large
task graphs.},
keywords = {Schedules, Processor scheduling, Evolutionary computation, Programming, Scheduling, Task analysis, Genetic algorithms, Processor scheduling, scheduling algorithms, genetic algorithms, optimization, mathematical programming, quadratic programming},
doi = {doi:10.1109/CEC45853.2021.9504867},
  notes = {Also known as \cite{9504867}},
)


@INPROCEEDINGS(Ma:2021:CEC,
%xplor 25 Aug 2021
  author = {Weijian Ma and Xinyuan Zhang and Yichen Xu and Fei Gao},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Loading is the Key: A Novel Genetic Quantum Algorithm for {SDVRP}},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1817--1825},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper solves Split Demand Vehicle Routing Problem with
minimal vehicles and controlled task splits. Our algorithm encodes the
mapping between task splits and vehicles into a binary matrix and uses
Genetic Quantum Algorithm to control the evolvement process. To convert
the binary matrix solution into task loading schemes, we design a novel
cost function and successfully convert task assignment problem to
Transportation Problem which can be solved by Transportation Simplex
Method. Our algorithm uses a simple nearest-neighborhood based heuristic
to generate vehicle routes and adopts a local search method tailored for
SDVRP to improve solution quality. The experimental results show that
our algorithm splits few tasks and can obtain many solutions better than
CVRP best-known in TSPLIB 95. Further analysis reveals that savings of
SDVRP mostly come from CVRP's failure to combine tasks geographically
close into one route, when the number of vehicles are restricted to
minimum.},
keywords = {Quantum algorithm, Loading, Vehicle routing, Transportation, Transforms, Genetics, Search problems, Split Demand Vehicle Routing Problem, Genetic Quantum Algorithm, Transportation Simplex Method},
doi = {doi:10.1109/CEC45853.2021.9504873},
  notes = {Also known as \cite{9504873}},
)


@INPROCEEDINGS(Acampora:2021:CEC,
%xplor 25 Aug 2021
  author = {Giovanni Acampora and Michele Grossi and Autilia Vitiello},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Algorithms for Error Mitigation in Quantum Measurement},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1826--1832},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Noisy Intermediate Scale Quantum (NISQ) devices are expected
to demonstrate the real potential of quantum computing in solving hard
problems. However, quantum noise that characterizes this kind of devices
still remains an obstacle for their practical exploitation in real world
scenarios. As a consequence, there is a strong emergence for error
correction techniques aimed at making NISQ devices stable and fully
operative. Unfortunately, current approaches for quantum error
correction are prohibitive for NISQ devices because of the enormous
multiplicative cost in resources that they require. For this reason,
so-called quantum error mitigation methods are emerging as alternative
approaches able to attenuate the quantum error as much as possible,
without requiring a strong additional computational effort. Among the
most error-prone operations, there is surely the quantum measurement.
Conventionally, mitigation methods for quantum measurement error compute
a so-called mitigation matrix capable of correcting results outputted by
a quantum processor. In this paper, a new measurement error mitigation
approach based on genetic algorithms is proposed to learn an appropriate
mitigation matrix. As shown in the experimental session, the proposed
measurement error mitigation method is comparable with or better than a
conventional algebraic approach in terms of the Hellinger fidelity.},
keywords = {Computers, Measurement errors, Quantum computing, Measurement uncertainty, Evolutionary computation, Genetics, Error correction, Quantum computing, Measurement error mitigation, Genetic algorithms},
doi = {doi:10.1109/CEC45853.2021.9504796},
  notes = {Also known as \cite{9504796}},
)


@INPROCEEDINGS(Saad:2021:CEC,
%xplor 25 Aug 2021
  author = {Hatem M.H. Saad and Ripon K. Chakrabortty and Saber Elsayed},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Quantum-Inspired Differential Evolution for Resource-Constrained
Project-Scheduling: Preliminary Study},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1833--1839},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Resource-Constrained Project Scheduling Problem (RCPSP) is
an NP-hard optimisation problem that can be found in many real-world
applications. Considerable research effort has been put into overcoming
the difficulties in solving the RCPSP by proposing innovative
heuristics, meta-heuristics and their hybridisation. However, finding
optimal solutions is still not guaranteed. It is known that
quantum-inspired metaheuristics can improve population diversity and the
quality of solutions but little has been published on adapting them to
solving RCPSPs. Here, we examine the performance of a Quantum-Inspired
Differential Evolution (QIDE) algorithm in solving such problems. The
proposed QIDE uses a quantum population that is initialised using the
rotation quantum gate and quantum superposition in the continuous
domain, and then evolved using the differential-evolution operators. A
local search is also adopted to accelerate convergence. The performance
of the QIDE algorithm was tested by solving problems with 30 and 60
activities from the PSPLIB benchmark datasets. The QIDE algorithm
outperformed another quantum-based particle swarm algorithm and some
other meta-heuristics.},
keywords = {Quantum computing, Processor scheduling, Sociology, Logic gates, Benchmark testing, Scheduling, Statistics, Quantum-Inspired Differential Evolution, quantum-inspired evolutionary algorithms, Resource-Constrained Project Scheduling},
doi = {doi:10.1109/CEC45853.2021.9504970},
  notes = {Also known as \cite{9504970}},
)


@INPROCEEDINGS(Antoniou:2021:CEC,
%xplor 25 Aug 2021
  author = {Margarita Antoniou and Gasper Petelin and Gregor Papa},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Preferred Solutions of the Ground Station Scheduling Problem
using NSGA-III with Weighted Reference Points Selection},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1840--1847},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Ground station Scheduling Problem refers to the allocation
of the communication tasks, among the ground stations and satellites. In
general, the problem is formulated as a many-objective problem. The
NSGA-III is an algorithm developed to solve such problems. Due to its
selection operator that uses a number of reference points, the NSGA-III
gives users the option to specify their own reference points. In this
paper, we use this opportunity by generating distributed reference
points, shifted depending on weights. A specific weight is assigned to
each objective that corresponds to reference points for preferred
solutions. The generation of these reference points and their effect on
the final objective function values of the Pareto front are first tested
on the DTLZ2 test function with 4 objectives and for a different
combination of weights. Finally, various weights are applied to an
instance of the Ground station Scheduling Problem, leading to Pareto
fronts that favor specific objectives and feasible schedules.},
keywords = {Schedules, Satellites, Processor scheduling, Evolutionary computation, Linear programming, Resource management, Task analysis, many-objective, satellite scheduling problem, NSGA-III, preferred solutions},
doi = {doi:10.1109/CEC45853.2021.9504886},
  notes = {Also known as \cite{9504886}},
)


@INPROCEEDINGS(Panda:2021:CEC,
%xplor 25 Aug 2021
  author = {Sai Panda and Yi Mei},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Programming with Algebraic Simplification for Dynamic Job
Shop Scheduling},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1848--1855},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Genetic Programming (GP) has been successfully applied to
evolve effective dispatching rules for Dynamic Job Shop Scheduling
(DJSS). However, the GP-evolved rules are usually too complex and hard
to interpret. In this paper, we aim to address this issue by evolving
simpler rules without losing effectiveness. To this end, we develop a
set of algebraic simplification operators based on our domain knowledge
about dynamic scheduling, which can recursively convert a rule into a
mathematically equivalent but simpler one. The algebraic simplification
operators can guarantee that the individual stays equivalent before and
after the simplification. Then, we develop a GP algorithm with these
simplification operators. We compared the GP with the simplification
operators with the baseline GP without simplification on a range of
scheduling instances, and the results showed that using the algebraic
simplification can slightly reduce the program size without sacrificing
the test performance of the evolved dispatching rules. Furthermore,
through deep analysis, we have also discovered the limitations of the
pure algebraic simplification for GP to evolve DJSS dispatching rules,
which can hardly simplify the individuals after the first generation.},
keywords = {Job shop scheduling, Heuristic algorithms, Genetic programming, Evolutionary computation, Dynamic scheduling, Dispatching, Dynamic programming},
doi = {doi:10.1109/CEC45853.2021.9505010},
  notes = {Also known as \cite{9505010}},
)


@INPROCEEDINGS(Cao:2021:CEC,
%xplor 25 Aug 2021
  author = {Yating Cao and Jing Liu and Zhouwu Xu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Dynamic Multiagent Genetic Algorithm for Optimal Charging in
Wireless Rechargeable Sensor Networks},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1856--1863},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Wireless energy charging is considered as a very promising
technology for prolonging sensor lifetime in Wireless Rechargeable
Sensor Networks (WRSNs). However, most existing studies suffers from
charging scalability and efficiency problems. Under such a background, a
new one to many charging model, which allows multiple sensors to be
charged simultaneously by a single mobile reader, is designed. A dynamic
multiagent genetic algorithm with a virtual force operator combined with
K-means clustering for WRSNs (D_VF_MAGA_WRSNs) is proposed to optimize
the charging problem in this paper. Agents representing candidate
solutions compete or cooperate with their neighbors, and can also use
knowledge. Memetic algorithm is applied in the process of evolution to
guide readers moving to the tags' clustering centers obtained by the
K-means algorithm. Moreover, a special crossover operator is designed to
dynamically adjust the number of reader charging positions in the
network. To verify the effectiveness of D_VF_MAGA_WRSNs, various of
experiments on different kinds of benchmarks are carried. Compared with
other algorithms, our algorithm is illustrated to be superior for WRSNs
in terms of total charging time, maximum charging load, charging
efficiency, and total charging distance.},
keywords = {Wireless communication, Wireless sensor networks, Heuristic algorithms, Scalability, Inductive charging, Dynamics, Force, Wireless rechargeable sensor networks, Multiagent genetic algorithm, Virtual force, Dynamic crossover operator, K-means algorithm},
doi = {doi:10.1109/CEC45853.2021.9504883},
  notes = {Also known as \cite{9504883}},
)


@INPROCEEDINGS(Cavalcanti-Costa:2021:CEC,
%xplor 25 Aug 2021
  author = {Joao Guilherme {Cavalcanti Costa} and Yi Mei and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Learning Initialisation Heuristic for Large Scale Vehicle Routing
Problem with Genetic Programming},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1864--1871},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Large Scale Vehicle Routing Problem is a classical NP-hard
problem. It has several applications in the industry and has always been
the focus of studies and development of new, ever more complex,
techniques to solve it. An important group of these techniques are Local
Search-based, which are sensitive to the initial solution given to them.
However, finding effective initial solutions is not a trivial task,
requiring domain knowledge for building them. Although some Genetic
Programming Hyper-Heuristics (GPHH) have tried to build better
heuristics automatically, they barely give an advantage for improving
the solution afterwards. This paper aims to show that Genetic
Programming can identify better regions of the search space, where the
initial solutions can be improved more efficiently with optimisation
steps. This is done by developing new terminals and a new fitness
function, which are based on the width of the routes, a metric that was
recently found to be an important feature for good solutions. The
obtained results show that the proposed approach finds better final
solutions than when using classical initial heuristics or other GPHH,
for both time efficiency and effectiveness.},
keywords = {Measurement, Space vehicles, Industries, NP-hard problem, Vehicle routing, Genetic programming, Search problems, Large Scale Vehicle Routing, Genetic Programming, Hyper-Heuristic, Initialisation},
doi = {doi:10.1109/CEC45853.2021.9504938},
  notes = {Also known as \cite{9504938}},
)


@INPROCEEDINGS(MacLachlan:2021:CEC,
%xplor 25 Aug 2021
  author = {Jordan MacLachlan and Yi Mei},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Look-Ahead Genetic Programming for Uncertain Capacitated Arc
Routing Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1872--1879},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Genetic Programming Hyper-Heuristic (GPHH) has been
successfully applied to evolve routing policies for the Uncertain
Capacitated Arc Routing Problem (UCARP). However, the current GPHH
approaches have a limitation that they only consider myopic information
of the current decision step. In this paper, we proposed incorporating
look-ahead information to the decision process of GP-evolved routing
policies. We designed a number of potentially promising chains of
candidate tasks, and expand the candidate task pool to consider both the
single tasks and task chains. This way, the routing policy can consider
the look-ahead information incorporated in the considered task chains.
The proposed GP with Chain Policies (GPCP) was compared with the
standard GPHH on a range of UCARP instances, and the results showed that
the task chains can improve the effectiveness of the routing policies
sometimes. The better performance of a routing policy largely depends on
whether it can balance the selections of single tasks and task chains,
and whether it can stick to the whole selected chain rather than only
the first task of the chain. In addition, there are some abnormal runs
with serious overfitting issue that we will address in our future work.},
keywords = {Genetic programming, Evolutionary computation, Routing, Task analysis, Standards},
doi = {doi:10.1109/CEC45853.2021.9504785},
  notes = {Also known as \cite{9504785}},
)


@INPROCEEDINGS(Yu:2021:CEC,
%xplor 25 Aug 2021
  author = {Yongbo Yu and Hui Ma and Gang Chen},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Achieving Multi-Objective Scheduling of Heterogeneous Workflows
in Cloud through a Genetic Programming Based Approach},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1880--1887},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Traditional human-designed heuristics-based algorithms are
commonly employed to address workflow scheduling problems. Such as
Heterogeneous Earliest Finish Time (HEFT) and CriticalPath are two
best-known list based heuristics. Generally, heuristics-based approaches
can generate a single heuristic by making decisions based on the current
status of the tasks and available resources and map the unscheduled
tasks to the available resources. However, traditional heuristics can
easily cause unbalancing load problem. For example, with the objective
of minimizing the makespan, traditional heuristics prefer to use
computation resources with high computation capacity. Usually, such
resources are expensive, which will lead to a high cost as a result.
Therefore it is hard for single traditional heuristics to do a
trade-off; thus, traditional heuristics are hard to apply for
multi-objective workflow scheduling. In this paper, we develop a novel
algorithm for workflow scheduling by considering both minimizing cost
and makespan. Experiments show that the MOSGP approach can effectively
minimize makespan and cost simultaneously.},
keywords = {Processor scheduling, Heuristic algorithms, Genetic programming, Evolutionary computation, Scheduling, Task analysis},
doi = {doi:10.1109/CEC45853.2021.9504695},
  notes = {Also known as \cite{9504695}},
)


@INPROCEEDINGS(Fan:2021:CEC,
%xplor 25 Aug 2021
  author = {Liang Fan and Xudong Feng and Handing Wang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Adaptive Formulation-based Many-Objective Evolutionary
Algorithm for Multi-Scenario Optimization in Data Enrichment},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1888--1895},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In many practical applications, data enrichment can generate a
large amount of accurate data to alleviate the problem of data scarcity.
In order to make the fake data generated in data enrichment as close to
the real data as possible, the data enriching model must be tuned to
meet the loss requirements of multiple objectives in different
scenarios, which makes it a multi-scenario many-objective optimization
problem. However, due to the curse of the dimensionality of the scenario
space and the objective space, the existing many-objective evolutionary
algorithms cannot solve the problem in data enrichment well. To
effectively handle this problem, we propose an adaptive
formulation-based multi-objective evolutionary algorithm, where the
aggregation function is used to reduce the dimension of the scenario
space to one and the multiple objectives into three objectives through
the adaptive formulation of the original problem. In this way, a
multi-scenario many-objective problem is converted into a
multi-objective problem which could be solved by existing
multi-objective evolutionary algorithms. The proposed algorithm is
applied to the practical data enrichment problem to solve the
multi-scenario many-objective optimization problem and compared with
NSGA-III. The experimental results demonstrate the remarkable
superiority of the proposed algorithm over NSGA-III.},
keywords = {Evolutionary computation, Data models, Optimization, Data enrichment, adaptive formulation, scenario reduction, multi-scenario optimization, many-objective optimization, multi-objective evolutionary algorithm},
doi = {doi:10.1109/CEC45853.2021.9504806},
  notes = {Also known as \cite{9504806}},
)


@INPROCEEDINGS(Fang:2021:CEC,
%xplor 25 Aug 2021
  author = {Junchao Fang and Wei Fang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Many-Objective Evolutionary Algorithm Based on New Angle
Penalized Distance},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1896--1903},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In evolutionary many-objective optimization, achieving better
balance between convergence and diversity of the population is a crucial
way to improve the efficiency of the algorithm. However, diversity
measure may select the individuals having good diversity but degrade the
convergence process to a certain extent. If the convergence measure
focuses on the convergence of the individuals too much, it may lead to
local convergence. The selection pressure achieves a severe loss,
especially when the Pareto dominance selection mechanism is difficult to
select solutions. To address these issues, a many-objective evolutionary
algorithm based on new angle penalized distance is proposed in this
paper, which is termed MaOEA-NAPD. In MaOEA-NAPD, it could dynamically
balance the convergence and diversity of the population concerning their
importance degree during the evolutionary process based on new angle
penalized distance. In order to enhance the selection probability of
better solutions in the mating pool, new convergence measure and
diversity measure are introduced according to the achievement
scalarizing function and angle based crowding degree estimation,
respectively. The performance of the proposed method is evaluated and
compared with five state-of-the-art algorithms on the WFG test suites
with up to 15 objectives. Experimental results show the superior
performance of MaOEA-NAPD than the compared algorithms on all the
considered test instances.},
keywords = {Heuristic algorithms, Sociology, Estimation, Evolutionary computation, Statistics, Optimization, Convergence, Many-objective optimization, Evolutionary algorithm, Angle penalized distance, Convergence, Diversity},
doi = {doi:10.1109/CEC45853.2021.9504935},
  notes = {Also known as \cite{9504935}},
)


@INPROCEEDINGS(Jerzy:2021:CEC,
%xplor 25 Aug 2021
  author = {Balicki Jerzy and Dryja Piotr},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Multi-objective Tabu-based Differential Evolution for
Teleportation of Smart Virtual Machines in Private Computing Clouds},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1904--1911},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We propose a multi-objective approach for using differential
evolution algorithm with tabu search algorithm as an additional mutation
for live migration (teleportation) of virtual machines. This issue is
crucial in private computing clouds. Teleportation of virtual machines
is supposed to be planned to determine Pareto-optimal solutions for
several criteria such as workload of the bottleneck host, communication
capacity of the critical node, electric power of hosts, and computer
costs. Some numerical results are presented for the experimental cloud
based on OpenStack platform for smart city and smart education systems.},
keywords = {Cloud computing, Smart cities, Evolutionary computation, Teleportation, Virtual machining, Power systems, Multi-objective Optimization, Live Migration of Virtual Machines, Private Computing Clouds},
doi = {doi:10.1109/CEC45853.2021.9504735},
  notes = {Also known as \cite{9504735}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {Longcan Chen and Lie Meng Pang and Hisao Ishibuchi and Ke Shang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Periodical Generation Update using an Unbounded External Archive
for Multi-Objective Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1912--1920},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In the evolutionary multi-objective optimization (EMO)
community, an unbounded external archive has been used in some studies
for evaluating the performance of EMO algorithms. Those studies show
that the unbounded external archive often includes better solutions than
the final population. Thus, it is likely that the search ability of an
EMO algorithm can be improved by periodically updating the current
population using the unbounded external archive (i.e., by periodically
choosing good solutions from all the examined solutions as the current
population). However, the usefulness of such a global generation update
scheme has not been studied in the literature. In this paper, we examine
the effect of the periodical global generation update on the performance
of well-known and frequently-used EMO algorithms: NSGA-II, MOEA/D and
NSGA-III. We use the PBI function with uniformly distributed weight
vectors for the periodical global generation update. In our
computational experiments, we obtain clearly improved results by the
periodical global generation update. We also examine the effect of the
frequency of the global generation update (e.g., every 20 generations)
on the performance of each EMO algorithm and its run time.},
keywords = {Time-frequency analysis, Sociology, Evolutionary computation, Search problems, Statistics, Optimization, Next generation networking, Evolutionary multi-objective optimization, many-objective optimization, global generation update, unbounded external archive, environmental selection},
doi = {doi:10.1109/CEC45853.2021.9504831},
  notes = {Also known as \cite{9504831}},
)


@INPROCEEDINGS(Venske:2021:CEC,
%xplor 25 Aug 2021
  author = {Sandra M. Venske and Carolina P. Almeida and Myriam R. Delgado},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Comparing Selection Hyper-Heuristics for Many-Objective Numerical
Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1921--1928},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Mechanisms for automatic selection of parameters/heuristics
used by evolutionary algorithms can provide more robust and independent
approaches. In this work we propose an approach composed of a selection
hyper-heuristic implemented within the MOEA/DD (Multi-objective
Evolutionary Algorithm based on Dominance and Decomposition) algorithm
based on Differential Evolution. Four selection hyper-heuristics are
considered in this study: Thompson Sampling, Probability Matching,
Adaptive Pursuit and Self-Adaptive Differential Evolution. The
hyper-heuristics are employed to choose the crossover operator selected
from a pool of operators, according to a probability that reflects the
operator's previous performance during the evolutionary process. The MaF
benchmark is considered with 5, 10 and 15 objectives. This benchmark
includes a diversity of characteristics, representing the challenges
that real-world problems may pose. Statistical tests indicate that the
proposed approach performs equally or even outperforms those with fixed
crossover operator.},
keywords = {Diversity reception, Evolutionary computation, Benchmark testing, Probability, Adaptive control, Optimization, Faces, Hyper-Heuristic, Many-Objective Optimization, MOEA/DD, MaFs},
doi = {doi:10.1109/CEC45853.2021.9504934},
  notes = {Also known as \cite{9504934}},
)


@INPROCEEDINGS(Wu:2021:CEC,
%xplor 25 Aug 2021
  author = {Kai Eivind Wu and George Panoutsos},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Visualisation Method for Pareto Front Approximations in
Many-objective Optimisation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1929--1937},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Visualisation of Pareto Front (PF) approximations of
many-objective optimisation problems (MaOP) is critical in understanding
and solving a MaOP. Research is ongoing on developing effective
visualisation methods with desired properties, such as simultaneously
revealing dominance relations, PF shape, and the diversity of
approximations. State-of-the-art visualisation methods in the literature
often retain some of the preferred properties, but there are still
shortfalls to address others. A new visualisation method is proposed in
this paper, which covers the majority of the desired properties for
visualisation methods. The proposed method is based on displaying PF
approximations via projections on a reference vector versus distances to
the same reference vector. The reference vector is created using nominal
Ideal and Nadir points of existing nondominated PF approximation sets.
MaF benchmark problems are used to demonstrate the effectiveness;
results show that the proposed method exhibits a more balanced
performance than the state-of-the-art in capturing desired visualisation
properties.},
keywords = {Visualization, Shape, Evolutionary computation, Benchmark testing, Optimization, Many-objective Optimisation, Performance indicator, Diversity, Reference vectors, Benchmark testing Introduction},
doi = {doi:10.1109/CEC45853.2021.9504904},
  notes = {Also known as \cite{9504904}},
)


@INPROCEEDINGS(Xia:2021:CEC,
%xplor 25 Aug 2021
  author = {Hai Xia and Changhe Li and Sanyou Zeng and Qingshan Tan and Junchen Wang and Shengxiang Yang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Reinforcement-Learning-Based Evolutionary Algorithm Using
Solution Space Clustering For Multimodal Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1938--1945},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In evolutionary algorithms, how to effectively select
interactive solutions for generating offspring is a challenging problem.
Though many operators are proposed, most of them select interactive
solutions (parents) randomly, having no specificity for the features of
landscapes in various problems. To address this issue, this paper
proposes a reinforcement-learning-based evolutionary algorithm to select
solutions within the approximated basin of attraction. In the algorithm,
the solution space is partitioned by the k-dimensional tree, and
features of subspaces are approximated with respect to two aspects:
objective values and uncertainties. Accordingly, two reinforcement
learning (RL) systems are constructed to determine where to search: the
objective-based RL exploits basins of attraction (clustered subspaces)
and the uncertainty-based RL explores subspaces that have been searched
comparatively less. Experiments are conducted on widely used benchmark
functions, demonstrating that the algorithm outperforms three other
popular multimodal optimization algorithms.},
keywords = {Uncertainty, Clustering algorithms, Evolutionary computation, Reinforcement learning, Benchmark testing, Search problems, Feature extraction, Evolutionary algorithm, reinforcement learning, landscape approximation, basin of attraction},
doi = {doi:10.1109/CEC45853.2021.9504896},
  notes = {Also known as \cite{9504896}},
)


@INPROCEEDINGS(Yang:2021:CEC,
%xplor 25 Aug 2021
  author = {Xu Yang and Rui Wang and Wenhua Li},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Helper Objective Assisted Evolutionary Algorithm for Multi-modal
Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1946--1952},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Evolutionary algorithms (EAs) have been widely applied in
various optimization problems. However, EAs are found as less effective
on multi-modal optimization due to their multiple local optima. Inspired
by the idea of self-paced learning, that is, problems can be solved step
by step, from easy to difficult. In this study, we propose to design a
helper objective function to assist the optimization of multi-modal
problems via multitask optimization framework. In principle the helper
objective function shares common features with the original function but
is easier to solve. Thus, the information gained by solving the helper
objective can be utilized to tackle the multi-modal problems.
Specifically, the Gaussian process is applied to build the helper
objective function. The multi-factorial evolutionary algorithm is
applied to optimize the helper and original objective functions
simultaneously. Experimental results show that the idea is effective on
a set of multi-modal optimization benchmarks.},
keywords = {Computational modeling, Evolutionary computation, Gaussian processes, Benchmark testing, Linear programming, Optimization, surrogate model, helper objective, multi-task optimization, Gaussian process modeling, multi-factorial evolutionary algorithm, differential evolution},
doi = {doi:10.1109/CEC45853.2021.9504932},
  notes = {Also known as \cite{9504932}},
)


@INPROCEEDINGS(Du:2021:CEC,
%xplor 25 Aug 2021
  author = {Wenhao Du and Zhigang Ren and An Chen and Hanqing Liu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Knowledge Transfer-Based Evolutionary Algorithm for Multimodal
Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1953--1960},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={There are some natural similarities among fitness landscapes
of different modals involved in a multimodal optimization problem
(MMOP). However, existing multimodal evolutionary algorithms (MMEAs)
tend to handle these modals separately, which limits their performance
to a great extent. As the first attempt, this study proposes a knowledge
transfer-based MMEA by mining and exploiting modal similarities. To this
end, a center-aligned normalization strategy (CANS) is first designed to
map all the subpopulations corresponding to identified modals into a
unified 0-1 space, and then a distribution similarity-based transfer
strategy (DSTS) is proposed to guide knowledge transfer among
subpopulations. DSTS explicitly quantifies the similarity between two
subpopulations and probabilistically transfers elite individuals in the
subpopulation of the largest similarity to the target subpopulation. By
integrating CANS and DSTS into a well-known MMEA, i.e., NEA2, the final
knowledge transfer-based NEA named KTNEA is developed. Experimental
results on CEC'2013 niching benchmark suite indicate that KTNEA performs
competitively in comparison with state-of-the-art MMEAs.},
keywords = {Shape, Evolutionary computation, Benchmark testing, Probabilistic logic, Optimization, Knowledge transfer, Multimodal optimization problem (MMOP), multimodal evolutionary algorithm (MMEA), knowledge transfer, NEA2},
doi = {doi:10.1109/CEC45853.2021.9504774},
  notes = {Also known as \cite{9504774}},
)


@INPROCEEDINGS(Luo:2021:CEC,
%xplor 25 Aug 2021
  author = {Wenjian Luo and Xin Lin and Jiajia Zhang and Mike Preuss},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Survey of Nearest-Better Clustering in Swarm and Evolutionary
Computation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1961--1967},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Nearest-Better Clustering (NBC) is an emergent niching
technique in Swarm and Evolutionary Computation for optimization, which
does not need to fix the number or radius of clusters in advance. The
key idea of NBC is to first link each individual to its nearest better
neighbor to form a spanning tree of all individuals in the population,
and then partition all individuals into clusters by deleting the longer
edges in the spanning tree. In this paper, a survey on the
Nearest-Better Clustering algorithms and applications in multimodal and
dynamic optimization is provided. First, the basic NBC algorithm is
introduced. Second, the improvements of the basic NBC are detailed.
Third, multimodal and dynamic optimization algorithms powered by NBC are
enlisted and discussed.},
keywords = {Heuristic algorithms, Clustering methods, Sociology, Clustering algorithms, Evolutionary computation, Vegetation, Partitioning algorithms, Evolutionary Algorithms, Swarm Intelligence, Nearest Better Clustering},
doi = {doi:10.1109/CEC45853.2021.9505008},
  notes = {Also known as \cite{9505008}},
)


@INPROCEEDINGS(Yang:2021:CEC,
%xplor 25 Aug 2021
  author = {Jian Yang and Yuhui Shi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Attention-oriented Brain Storm Optimization for Multimodal
Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1968--1975},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Population-based methods are often used to solve multimodal
optimization problems. By combining niching or clustering strategy, the
state-of-the-art approaches generally divide the population into several
subpopulations to find multiple solutions for a problem at hand.
However, these methods only guided by the fitness value during
iterations, which are suffering from determining the number of
subpopulations, i.e., the number of niche areas or clusters. To
compensate for this drawback, this paper presents an Attention-oriented
Brain Storm Optimization (ABSO) method that introduces the attention
mechanism into a relatively new swarm intelligence algorithm, i.e.,
Brain Storm Optimization (BSO). By converting the objective space from
the fitness space into "attention" space, the individuals are clustered
and updated iteratively according to their salient values. Rather than
converge to a single global optimum, the proposed method can guide the
search procedure to converge to multiple "salient" solutions. The
preliminary results show that the proposed method can locate multiple
global and local optimal solutions of several multimodal benchmark
functions. The proposed method needs less prior knowledge of the problem
and can automatically converge to multiple optimums guided by the
attention mechanism, which has excellent potential for further
development.},
keywords = {Storms, Sociology, Clustering algorithms, Benchmark testing, Linear programming, Extraterrestrial measurements, Statistics, Multimodal Optimization, Brain Storm Optimization, Attention Mechanism, Population-based Algorithms},
doi = {doi:10.1109/CEC45853.2021.9504871},
  notes = {Also known as \cite{9504871}},
)


@INPROCEEDINGS(Li:2021:CEC,
%xplor 25 Aug 2021
  author = {Dongyang Li and Weian Guo and Lei Wang and Qidi Wu},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Modified {APSODEE} for Large Scale Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1976--1982},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The balance of exploration and exploitation of particle swarm
optimization (PSO) is still a great challenge in large scale
optimization. In our previous work, an adaptive particle swarm optimizer
with decoupled exploration and exploitation (APSODEE) has been proposed
to solve this problem. However, it still shows room for further
improvements. First, APSODEE guides particles with the best individual
in each sub-swarm, which is adverse to swarm diversity preservation
especially in case of large scale optimization. Second, APSODEE fails to
lead particles moving to sparse areas which are also potentially
promising. To address these issues, two modifications are incorporated
into APSODEE including a partial updating strategy and a quality
restrained local sparseness diversity measurement. The former is
proposed to further enhance the algorithm's swarm diversity preservation
ability while the latter is designed to help guiding updated particles
moving towards the areas which are both sparse and potentially
promising, resulting in the modified APSODEE. The experiments are
conducted based on CEC 2013 benchmarks with 1000 dimensionality. The
results show the competitiveness of the proposed algorithm.},
keywords = {Atmospheric measurements, Benchmark testing, Particle measurements, Particle swarm optimization, Optimization, particle swarm optimization, balance of exploration and exploitation, large scale optimization, partial updating strategy, quality restrained local sparseness diversity},
doi = {doi:10.1109/CEC45853.2021.9505003},
  notes = {Also known as \cite{9505003}},
)


@INPROCEEDINGS(Vodopija:2021:CEC,
%xplor 25 Aug 2021
  author = {Aljosa Vodopija and Beate Breiderhoff and Boris Naujoks and Bogdan Filipic},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Design of Cyclone Dust Separators: A Constrained Multiobjective
Optimization Perspective},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1983--1990},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Cyclone dust separators (cyclones for short) are devices used
to remove dispersed particles from a flue gas. They are widely applied
because of their simple structure, low cost, ease of operation, and
ability to operate in harsh environments. In contrast to previous work
where the cyclone dust separator design was mainly regarded as an
optimization problem with box constraints, we approach it as an
optimization task involving the maximization of the collection
efficiency and the minimization of the pressure drop, while regarding
both box constraints and geometric constraints that limit the designs.
We study three highefficiency cyclone configurations, define four test
instances for each configuration, and analyze the feasibility ratios
resulting from the constraints. In addition, we investigate the
performance of three multiobjective optimization algorithms on this
design problem. We analyze the results, both in the objective and the
decision space, investigate the effect of the decision space size on the
solution quality, and gain insights into problem properties relevant to
cyclone designers. Finally, we illustrate the results with selected
optimized cyclone designs.},
keywords = {Constraint handling, Visualization, Handheld computers, Particle separators, Evolutionary computation, Color, Cyclones, engineering design, cyclone dust separator, collection efficiency, pressure drop, multiobjective optimization},
doi = {doi:10.1109/CEC45853.2021.9504991},
  notes = {Also known as \cite{9504991}},
)


@INPROCEEDINGS(Grin:2021:CEC,
%xplor 25 Aug 2021
  author = {Aleksandr V Grin and Amir H Gandomi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Advancing Genetic Programming via Information Theory},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1991--1998},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Genetic Programming (GP) is a powerful tool often used to
solve optimization problems where analytical methods are unusable. While
the general technique is well understood, there exist deficiencies in
the multitude of implementations currently widely available. The primary
areas of improvement are computation time, search space reduction, and
accuracy. Despite significant advances in GP systems, a key deficiency
remains in the structural randomization of symbolic GP trees. Our
initial assumptions regarding the formation of expression trees in
symbolic GP trees is at best highly limited and normally simply
non-existent. In this paper, we introduce a new GP methodology that
incorporates both current cutting- edge GP system solutions as well as
an information-theoretic approach to expression tree initialization.
Through a more informed initial tree construction, this approach reduces
the search space and model complexity. We introduce in this work the
methodology as well as the accompanying theoretical component and
comparison benchmarks from tests. A key advantage of the algorithm
proposed is its high parallelization potential which is highlighted in
further discussion. The method consists of two parts. The first is a
variable-interaction system termed Entropy Shaving that is used for both
variable selection and initial expression structure generation. The
second is a GP system that utilizes the variable-interaction system as
input to determine a final solution.},
keywords = {Systematics, Input variables, Genetic programming, Evolutionary computation, Tools, Search problems, Entropy, Genetic Programming, Information Theory, Data Analytics, Evolutionary Computation},
doi = {doi:10.1109/CEC45853.2021.9504859},
  notes = {Also known as \cite{9504859}},
)


@INPROCEEDINGS(Mohebali:2021:CEC,
%xplor 25 Aug 2021
  author = {Behshad Mohebali and Gelareh Karbaschi and Amirhessam Tahmassebi and Anke Meyer-Baese and Amir H. Gandomi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Evolutionary Framework for Real-Time Fraudulent Credit
Detection},
year={2021},
  editor = {Yew-Soon Ong},
pages = {1999--2006},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Fraud has been a worldwide issue that is facing the major
economies of the world. Within an economical system, undetected and
unpunished fraudulent activities can erode the public trust in law
enforcement institutions and even incentivize more fraud. Therefore,
detection of fraudulent activities and prosecution of responsible
entities is of utmost importance for financial regulatory bodies around
the globe. Of the challenges rising with this task is the scarcity of
detection resources (auditors) and the fraudsters constantly adapting to
the new circumstances of the market. To address these issues, this paper
proposes an evolutionary framework for credit fraud detection with the
ability to incorporate (and adapt to) the incoming data in real-time.
The goal of the framework is to identify the entities with high a risk
of fraud for efficient targeting of the scarce resources. The data that
is generated as a result of the audits are fed into the framework for
further training.},
keywords = {Training, Law enforcement, Evolutionary computation, Real-time systems, Task analysis},
doi = {doi:10.1109/CEC45853.2021.9504729},
  notes = {Also known as \cite{9504729}},
)


@INPROCEEDINGS(Souto:2021:CEC,
%xplor 25 Aug 2021
  author = {Gabriel Souto and Igor Morais and Liss Faulhaber and Glaydston Mattos Ribeiro and Pedro {Henrique Gonzalez}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Hybrid {BRKGA} Approach for the Two Stage Capacitated Facility
Location Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2007--2014},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Finding optimal locations for installing factories and
warehouses, fulfilling customers' demand is the goal of the Two-Stage
Capacitated Facility Location, which compounds a supply chain class of
problems. Besides locating factories and warehouses, the problem's
objective is to minimize the operational costs: opening facilities and
product flow, satisfying customer demand, and without disregarding the
capacity of factories and warehouses. A Hybrid Biased Random-Key Genetic
Algorithm (HBRKGA), combining BRKGA with Local Search and Local
Branching, was proposed to solve the problem. The computational
experiments showed that the method was reliable and stood out among
other approaches to the problem.},
keywords = {Supply chains, Evolutionary computation, Search problems, Production facilities, Reliability, Compounds, Genetic algorithms, Two-Stage Capacitated Facility Location, BRKGA, Matheuristic, Industries 4.0, Supply Chain},
doi = {doi:10.1109/CEC45853.2021.9504856},
  notes = {Also known as \cite{9504856}},
)


@INPROCEEDINGS(Yan:2021:CEC,
%xplor 25 Aug 2021
  author = {Zichu Yan and Ying Bi and Bing Xue and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Automatically Extracting Features Using Genetic Programming for
Low-Quality Fish Image Classification},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2015--2022},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Fish image classification is an important task in the
protection of precious marine resources. However, this task is difficult
due to the low-quality images and the high inter-class variations across
images. Most existing methods use high-quality images for classification
and need domain knowledge. In this paper, we develop a genetic
programming (GP) approach to automatically selecting image operators to
deal with the low-quality images and extracting effective features from
these images for low-quality fish image classification. To achieve this,
a new program structure and a new function set are developed. With these
designs, the proposed GP approach can evolve solutions that use
effective filtering or restoration operators to deal with the input
image, select informative regions from the fish image, and extract
effective global and/or local features from the fish images. The results
show that the proposed approach achieves significantly better
performance than 12 benchmark methods, including a state-of-the-art GP
approach, on the well-known fish image classification dataset. Further
analysis shows the high interpretability of the evolved GP trees and the
effectiveness of the employed image filtering or restoration operators.},
keywords = {Filtering, Genetic programming, Evolutionary computation, Benchmark testing, Feature extraction, Fish, Image filtering},
doi = {doi:10.1109/CEC45853.2021.9504737},
  notes = {Also known as \cite{9504737}},
)


@INPROCEEDINGS(Chang:2021:CEC,
%xplor 25 Aug 2021
  author = {Chun-yien Chang and Ying-ping Chen},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Contrapuntal Composition and Autonomous Style Development of
Organum Motets by using AntsOMG},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2023--2030},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Based on a previously proposed meta-framework, called ants on
multiple graphs (AntsOMG), this paper further investigates into certain
more complicated creative behavior, in which creators interact with
their own creations in order to complete works and elevate the creation
forms to a higher level, via computational mechanisms. Aiming to achieve
the goal of this study, the target music genre, organum motets, is
adopted because its composition process requires iterative interactions
with composed music and quite resembles the creative behavior under
investigation. AntsOMG is employed as a fundamental component for
developing models and conducting creations according to the developed
models. With the assistance of AntsOMG, the proposed approach firstly
develops style models for section scheme planning and accordingly plans
section schemes for the organum motets to be composed. Secondly,
following the section scheme, cantus firmus is composed and also used to
generate the graph models to be developed into the style models required
for composing the corresponding contrapuntal parts for the polyphonic
sections of the organum motets. For finalizing each contrapuntal
section, a genetic algorithm is utilized to introduce variety and
diversity. The outcomes demonstrate that contrapuntal composition and
autonomous style development of organum motets can be achieved by the
proposed approach. The contribution of this paper is twofold. First, the
presented implementation is immediately applicable to compose unlimited
amount of organum motets, which may not be possible for human composers.
Second, the success of the proposal may shed light on gaining further
understandings of complicated creative behavior.},
keywords = {Computational modeling, Transforms, Evolutionary computation, Planning, Proposals, Iterative methods, Guidelines},
doi = {doi:10.1109/CEC45853.2021.9504881},
  notes = {Also known as \cite{9504881}},
)


@INPROCEEDINGS(Zheng:2021:CEC,
%xplor 25 Aug 2021
  author = {Wei Zheng and Xin Liu and Jianyong Sun},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Continuous Encoding for Community Detection in Attribute Networks
with Preserving Node Information},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2031--2038},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Community detection in complex attribute network is an
indispensable but difficult task in data mining. Recently, using
multiobjective evolutionary algorithm (MOEA) to address this task has
become popular since it can be naturally modeled as a discrete
multiobjective optimization problem (MOP). In this paper, we develop a
continuous MOEA, in which a continuous encoding is proposed to convert
the discrete MOP into a continuous one by introducing a set of auxiliary
continuous variables. Further, we construct a similarity matrix to
replace the adjacency matrix by making use of the network node degree
information in the encoding. The new similarity matrix not only reserves
the property of the adjacency matrix but includes the degree information
of all the network nodes. In our experiments, various benchmark networks
with or without ground truths are used to compare with some
state-of-the-art MOEA-based and non-MOEA-based methods. The experimental
results show that the proposed algorithm performs favorably against the
compared methods.},
keywords = {Measurement, Social networking (online), Evolutionary computation, Complex networks, Encoding, Data mining, Matrix converters, Complex attribute network, Community detection, Node information, Similarity matrix, Multi-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504842},
  notes = {Also known as \cite{9504842}},
)


@INPROCEEDINGS(Sainz-Tinajero:2021:CEC,
%xplor 25 Aug 2021
  author = {Benjamin Mario Sainz-Tinajero and Andres Eduardo Gutierrez-Rodriguez and Hector G. Ceballos and Francisco J. Cantu-Ortiz},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolutionary Clustering Algorithm Using Supervised Classifiers},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2039--2045},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Cluster analysis is a data mining tool for searching patterns
automatically on different types of data. However, it is not always
clear which clustering criterion would be the most accurate as this
decision is domain-dependent. This paper focuses on the design of a
single-objective evolutionary clustering algorithm that generates
solutions that are less biased towards one cluster structure. This
work's motivation starts from the idea that a good partition induces a
well-trained classifier. The resulting evolutionary clustering algorithm
using classifiers aims to enhance the generalization capability of the
resulting partition. The proposed objective function trains a set of
classifiers using an individual's chromosome as class labels. The
obtained average area under the curve of the classifiers is used as the
cluster quality index for measuring fitness. The experimental phase
shows a similarity increase between the generated solutions and the
reference partitions of the datasets.},
keywords = {Scalability, Clustering algorithms, Evolutionary computation, Tools, Linear programming, Search problems, Classification algorithms, data mining, unsupervised learning, clustering, evolutionary clustering, optimization},
doi = {doi:10.1109/CEC45853.2021.9504826},
  notes = {Also known as \cite{9504826}},
)


@INPROCEEDINGS(Brezinski:2021:CEC,
%xplor 25 Aug 2021
  author = {Kenneth Brezinski and Ken Ferens},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Complexity-Based Lambda Layer for Time Series Prediction},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2046--2052},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Time series analysis forms the basis for the temporal
sequences that we observe in everyday natural phenomenon. Examining and
characterizing time series' forms the basis for research in areas
spanning classifying heart rate variability, temperature prediction and
stock price prediction. In recent history combining powerful techniques
such as Fourier Analysis with Machine Learning techniques has improved
our ability to predict future time series based on historic data.
Complexity is one of such tools that incorporates long-range dependency
based on the inherent self-similarity that exists in many natural
phenomena. Leveraging the prowess of recurrent neural networks with that
of complexity measures combines two very powerful techniques to improve
prediction accuracy. In this work a Complexity Lambda Layer was
initialized in series with Artificial Neural Networks (ANN)
architectures to improve prediction accuracy for synthesized Brownian
Noise. Window size for recurrence and the stationary interval size was
optimized for increased performance. For both non-temporal and temporal
studies, a 2-4 fold improvement in root-mean-squared accuracy was
obtained. This approach was implemented as a Lambda layer, meaning the
improvement can be done on-the fly with minimal overhead.},
keywords = {Training, Recurrent neural networks, Time series analysis, Artificial neural networks, Computer architecture, Machine learning, Tools, Complexity, Recurrent Neural Networks, Time Series, Machine Learning},
doi = {doi:10.1109/CEC45853.2021.9504995},
  notes = {Also known as \cite{9504995}},
)


@INPROCEEDINGS(Wang:2021:CEC,
%xplor 25 Aug 2021
  author = {Peng Wang and Bing Xue and Mengjie Zhang and Jing Liang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Grid-dominance based Multi-objective Algorithm for Feature
Selection in Classification},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2053--2060},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Feature selection aims to select a small subset of relevant
features while maintaining or even improving the classification
performance over using all features. Feature selection can be considered
as a multi-objective problem, i.e., minimizing the number of selected
features and maximizing the classification accuracy (minimizing the
classification error) simultaneously. Most evolutionary multi-objective
algorithms encounter difficulties when handling a feature selection task
due to the discrete search space, although they perform well on
continuous/numeric optimization problems. This paper proposes a
grid-dominance based multi-objective evolutionary algorithm to address
feature selection. The aim is to explore the potential of the
grid-dominance method to strengthen the selection pressure toward the
optimal direction while maintaining an extensive distribution among the
objective values of feature subsets. To increase the population
diversity, a subset filtration mechanism is proposed. The performance of
the proposed two algorithms is tested on fourteen datasets of varying
difficulty. With the proposed methods, the performance metrics,
hypervolume and inverted generational distance have been significantly
improved compared with other commonly used multi-objective algorithms,
and the population diversity has also been increased.},
keywords = {Measurement, Filtration, Sociology, Evolutionary computation, Feature extraction, Search problems, Classification algorithms},
doi = {doi:10.1109/CEC45853.2021.9504832},
  notes = {Also known as \cite{9504832}},
)


@INPROCEEDINGS(Smith:2021:CEC,
%xplor 25 Aug 2021
  author = {Robert J. Smith and Ryan Amaral and Malcolm I. Heywood},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolving Simple Solutions to the {CIFAR-10} Benchmark using Tangled
Program Graphs},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2061--2068},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The goal of the CIFAR-10 benchmark is recast from the
perspective of discovering light-weight as well as accurate solutions.
Specifically, the image data, on which CIFAR-10 is based, requires
multiple practical issues to be addressed that are not often considered
collectively when applying genetic programming to classification
problems. Issues of particular interest include cardinality, multi-class
classification and diversity maintenance. We demonstrate that diversity
maintenance and cardinality can be approached simultaneously by adopting
a data subset to compose pools of exemplars for lexicase selection. The
issues of multi-class classification and solution simplicity are
addressed by adopting the tangled program graph (TPG) approach to
emergent modularity. In addition, the mutation operator is modified to
ensure that class labels do not 'die out' during evolution. The
resulting benchmarking study demonstrates solutions that are
significantly more accurate than AutoML while providing comparable
accuracies with solutions from unsupervised feature discovery, i.e. 7percent
accuracy. However, unlike the latter TPG solutions are several orders of
magnitude simpler.},
keywords = {Training, Sociology, Supervised learning, Genetic programming, Training data, Lighting, Benchmark testing, CIFAR-10, Tangled program graphs, Modularity, Lexicase Selection, k-armed Bandit},
doi = {doi:10.1109/CEC45853.2021.9504998},
  notes = {Also known as \cite{9504998}},
)


@INPROCEEDINGS(Klopotek:2021:CEC,
%xplor 25 Aug 2021
  author = {Mieczyslaw A. Klopotek and Slawomir T. Wierzchon},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Hullingversus Clustering - Two Complementary Applications of
Non-Negative Matrix Factorization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2069--2076},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this paper we make a comparison of two NMF based techniques
of dataset characterization: clustering and hulling. The characteristics
of a dataset should be understood as describing the content of a data
set through several characteristic representatives. Hulling (defined
later) characterizes the data by saying that the data points are
somewhere between the representatives, while clustering characterizes
the data by saying that the data points are close to one or the other
representative. The precision of such a characteristic will be measured
as a deviation from the idea of characterization, i.e. the distance of
the actual data points from the closest representatives in the case of
clustering and from the interior of the hull spanned by the
representatives. We show that for low-dimensional data the hull-based
characterization precision is much better than in case of clustering.
Clustering and hulling are two examples of sophisticated optimization
problems. Evolutionary algorithms are an excellent tool for solving such
problems. However, in the case of large, high-dimensional data sets,
their usefulness decreases. In this paper, we discuss heuristics for
hulling for massive data. We hope that it will inspire the creation of
an effective evolutionary algorithm dedicated to solving such problems.},
keywords = {Three-dimensional displays, Evolutionary computation, Tools, Generators, Proposals, Matrix decomposition, Optimization, Non-negative Matrix Factorization, clustering, hulling, k-means},
doi = {doi:10.1109/CEC45853.2021.9504734},
  notes = {Also known as \cite{9504734}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {Brian Chen and Siddhant Tandon and David Gorsich and Alex Gorodetsky and Shravan Veerapaneni},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Behavioral Cloning in Atari Games Using a Combined Variational
Autoencoder and Predictor Model},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2077--2084},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We explore an approach to behavioral cloning in video games.
We are motivated to pursue a learning architecture that is data
efficient and provides opportunity for interpreting player strategies
and replicating player actions in unseen situations. To this end, we
have developed a generative model that learns latent features of a game
that can be used for training an action predictor. Specifically, our
architecture combines a Variational Autoencoder with a discriminator
mapping the latent space to action predictions (predictor). We compare
our model performance to two different behavior cloning architectures: a
discriminative model (a Convolutional Neural Network) mapping game
states directly to actions, and a Variational Autoencoder with a
predictor trained separately. Finally, we demonstrate how we can use the
advantage of generative modeling to sample new states from the latent
space of the Variational Autoencoder to analyze player actions and
provide meaning to certain latent features.},
keywords = {Training, Computational modeling, Cloning, Games, Evolutionary computation, Computer architecture, Predictive models, Index Terms--Behavior Cloning, Variational Autoencoder, Predictive Models, Video Games},
doi = {doi:10.1109/CEC45853.2021.9505001},
  notes = {Also known as \cite{9505001}},
)


@INPROCEEDINGS(Dockhorn:2021:CEC,
%xplor 25 Aug 2021
  author = {Alexander Dockhorn and Jorge Hurtado-Grueso and Dominik Jeurissen and Linjie Xu and Diego Perez-Liebana},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Portfolio Search and Optimization for General Strategy
Game-Playing},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2085--2092},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Portfolio methods represent a simple but efficient type of
action abstraction which has shown to improve the performance of
search-based agents in a range of strategy games. We first review
existing portfolio techniques and propose a new algorithm for
optimization and action-selection based on the Rolling Horizon
Evolutionary Algorithm. Moreover, a series of variants are developed to
solve problems in different aspects. We further analyze the performance
of discussed agents in a general strategy game-playing task. For this
purpose, we run experiments on three different game-modes of the
Stratega framework. For the optimization of the agents' parameters and
portfolio sets we study the use of the N-tuple Bandit Evolutionary
Algorithm. The resulting portfolio sets suggest a high diversity in
play-styles while being able to consistently beat the sample agents. An
analysis of the agents' performance shows that the proposed algorithm
generalizes well to all game-modes and is able to outperform other
portfolio methods.},
keywords = {Economics, Optimization methods, Evolutionary computation, Games, Space exploration, Task analysis, Portfolios, Portfolio Methods, General Strategy Game-playing, Stratega, N-Tuple Bandit Evolutionary Algorithm},
doi = {doi:10.1109/CEC45853.2021.9504824},
  notes = {Also known as \cite{9504824}},
)


@INPROCEEDINGS(Flimmel:2021:CEC,
%xplor 25 Aug 2021
  author = {Julius Flimmel and Jakub Gemrot and Vojtech Cerny},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Coevolution of {AI} and Level Generators for Super Mario Game},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2093--2100},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Procedural content generation (PCG) is now used in many games
to generate a wide variety of content. One way of evaluating this
content is by artificial intelligence (AI) controlled players.
Inversely, PCG content can also be used when training AI players to
ensure generalization. Evolutionary algorithms are employed in both AI
and PCG fields, but rarely simultaneously. In this work, we use
evolutionary algorithms for both AI players and level generation in the
platformer game Super Mario. We further combine them into a coevolution,
where the AI players are evaluated by adapting level generators, and
vice versa, level generators are evaluated by adapting AI players. This
yields an AI player trained on gradually more difficult levels and a
sequence of level generators with gradually increasing difficulty. Such
sequence of generators might be useful for human game playing in
commercial games.},
keywords = {Measurement, Training, Level set, Linearity, Games, Evolutionary computation, Markov processes, procedural content generation, artificial intelligence, coevolution, evolutionary algorithms, Super Mario},
doi = {doi:10.1109/CEC45853.2021.9504742},
  notes = {Also known as \cite{9504742}},
)


@INPROCEEDINGS(Shakya:2021:CEC,
%xplor 25 Aug 2021
  author = {Siddhartha Shakya and Kin Poon and Khawla AlShanqiti and Anis Ouali and Andrei Sleptchenko},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Investigating binary {EAs} for Passive In-Building Distributed
Antenna Systems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2101--2108},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A passive in-building distributed antenna system (IB-DAS) is
often used to enhance indoor mobile data coverage by introducing indoor
antennas inside buildings. Such systems are created to ensure that
traffic generated indoors does not heavily depend on base stations
installed outdoor, as penetration issues of wireless signals can affect
the quality of connection. The focus of this paper is on the automation
of IB-DAS design. Particularly, it provides an extensive analysis of the
performance of four binary Evolutionary Algorithms (EAs) for this
problem and shows that the two tested Estimation of Distribution
Algorithms (EDAs) performs well on this problem. Furthermore, it
investigates the effect of different genetic operators on the
performance of the considered EAs. The practice outcome of this is to
select the best algorithm among others to be implemented in a DAS
network planning tool and to help our industrial partner reduce both the
design time and deployment cost.},
keywords = {Computational modeling, Estimation, Evolutionary computation, Tools, Mobile antennas, Complexity theory, Planning, Distributed Antenna System, Evolutionary Algorithm, GA, EDA, splitters, antennas},
doi = {doi:10.1109/CEC45853.2021.9504731},
  notes = {Also known as \cite{9504731}},
)


@INPROCEEDINGS(Cavalcanti-Costa:2021:CEC,
%xplor 25 Aug 2021
  author = {Joao Guilherme {Cavalcanti Costa} and Yi Mei and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Evolutionary Hyper-Heuristic Approach to the Large Scale
Vehicle Routing Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2109--2116},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Large Scale Vehicle Routing Problem (LSVRP) is a classical
combinatorial optimisation problem that serves several customers on a
graph using a set of vehicles. Due to the NP-hardness and large problem
size, LSVRP cannot be efficiently solved by exact approaches. Heuristic
methods such as the Iterative Local Search or the Hybrid Genetic
Algorithm still struggle for finding effective solutions for large scale
instances. For these methods to deal with the large search space,
pruning techniques are applied in order to limit the number of explored
solutions. However, effective pruning is a hard task, requiring domain
knowledge to craft good ways of limiting the search space without losing
the ability to find better solutions. Hyper-heuristics are types of
methods that aim to reduce domain knowledge on the creation of
heuristics, and in this work, we also apply them for effective heuristic
pruning. Our Evolutionary Hyper-Heuristic (EHH) automatically evolves
limits to the solution search space together with the heuristic utilised
to build and improve solutions for the LSVRP. We utilise a Guided Local
Search (GLS) as the base algorithm in which our EHH searches for the
best heuristic configuration. Our results show that the EHH can find
better solutions for most LSVRP test instances when compared to the
manually designed pruning of the GLS.},
keywords = {Training, Limiting, Heuristic algorithms, Vehicle routing, Evolutionary computation, Search problems, Iterative algorithms, Vehicle Routing Problem, Large Scale, Hyper-Heuristics, Automatic Pruning},
doi = {doi:10.1109/CEC45853.2021.9504818},
  notes = {Also known as \cite{9504818}},
)


@INPROCEEDINGS(Xu:2021:CEC,
%xplor 25 Aug 2021
  author = {Meng Xu and Fangfang Zhang and Yi Mei and Mengjie Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Genetic Programming with Archive for Dynamic Flexible Job Shop
Scheduling},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2117--2124},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Genetic programming (GP) has achieved great success in
evolving effective scheduling rules to make real-time decisions in
dynamic flexible job shop scheduling (DFJSS). To improve generalization,
a commonly used strategy is to change the training simulation(s) at each
generation of the GP process. However, with such a simulation rotation,
GP may lose potentially promising individuals that happen to perform
poorly in one particular generation. To address this issue, this paper
proposed a new multi-tree GP with archive (MTAGP) to evolve the routing
and sequencing rules for DFJSS. The archive is used to store the
potentially promising individuals of each generation during evolution of
genetic programming. The individuals in the archive can then be fully
utilized when the simulation is changed in subsequent generations.
Through extensive experimental tests, the MTAGP algorithm proposed in
this paper is more effective than the multi-tree GP without archive
algorithm in a few scenarios. Further experiments were carried out to
analyze the use of the archive and some possible guesses were ruled out.
We argue that the use of archives does increase the diversity of the
population. However, the number of individuals in the archive that
ranked in the top five of the new population is small. Therefore, the
archive may not be able to greatly improve the performance. In the
future, we will investigate better ways to use the archive and better
ways to update individuals in the archive.},
keywords = {Training, Sequential analysis, Job shop scheduling, Sociology, Genetic programming, Dynamic scheduling, Routing, dynamic flexible job shop scheduling, genetic programming, archive},
doi = {doi:10.1109/CEC45853.2021.9504752},
  notes = {Also known as \cite{9504752}},
)


@INPROCEEDINGS(Planinic:2021:CEC,
%xplor 25 Aug 2021
  author = {Lucija Planinic and Marko ?urasevic and Domagoj Jakobovic},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={On the Application of ?-Lexicase Selection in the Generation of
Dispatching Rules},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2125--2132},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Dynamic online scheduling is a difficult problem which
commonly appears in the real world. This is because the decisions have
to be performed in a small amount of time using only currently available
incomplete information. In such cases dispatching rules (DRs) are the
most commonly used methods. Since designing them manually is a difficult
task, this process has been successfully automatised by using genetic
programming (GP). The quality of the evolved rules depends on the
problem instances that are used during the training process. Previous
studies demonstrated that careful selection of problem instances on
which the solutions should be evaluated during evolution improves the
performance of the generated rules. This paper examines the application
of the ?-lexicase selection to the design of DRs for the unrelated
machines scheduling. This selection offers a better solution diversity
since the individuals are selected based on a smaller subset of
instances, which leads to the creation of DRs that perform well on the
selected instances. The experiments demonstrate that this type of
selection can significantly improve the results for the Roulette Wheel
and Elimination GP variants, while achieving the same performance as the
Steady State Tournament GP. Furthermore, the ?-lexicase based algorithms
have a better convergence rate, which means that the increased diversity
in the population has a positive effect on the evolution process.},
keywords = {Training, Sociology, Wheels, Genetic programming, Dynamic scheduling, Dispatching, Steady-state, lexicase, genetic programming, dispatching rules, scheduling},
doi = {doi:10.1109/CEC45853.2021.9504982},
  notes = {Also known as \cite{9504982}},
)


@INPROCEEDINGS(Portela:2021:CEC,
%xplor 25 Aug 2021
  author = {Braulio M.O. Portela and Heder S. Bernardino and Luciana B. Goncalves and Stenio S. R. F. Soares},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Cheapest Insertion and Disruption of Routes Operators for Solving
Multi-Depot Electric Vehicle Location Routing Problem with Time Windows
and Battery Swapping via GRASP and RVND},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2133--2140},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Multi-Depot Electric Vehicle Location Routing Problem with
Time Windows and Battery Swapping (MDEVLRPTW-BS) aims to mitigate the
environmental impacts caused by the use of vehicles with an internal
combustion engine and uses economically viable electric vehicles in
transportation services with time window restrictions. This problem is
complex as one must define (i) the location of the battery swapping
stations, (ii) the depots to be used from a list of candidates, and
(iii) the assignment and routing of a fleet of electric vehicles. We
propose here two constructive methods to solve MDEVLRPTW-BS, namely,
Cheapest Insertion and Disruption of Routes. These constructive
approaches are used with the Greedy Randomized Adaptive Search Procedure
(GRASP) and the Random Variable Neighborhood Descent (RVND) method with
six movement operators. Two of these operators, namely Isolate and
ChangeBSS, are also proposed here for solving MDEVLRPTW-BS. The
proposals are compared with a technique from the literature using
instances with 5, 10, and 15 customers, and the proposed approaches
presented good results, mainly regarding the processing time. Also, we
propose new sets of instances larger than those from the literature.
Thus, the search techniques can be analyzed when subjected to greater
and more realistic situations. These new instances are composed of 100,
144, 288, 360, 420, and 600 customers. For these cases, the approaches
also obtained good and promising results, mainly when the Disruption of
Routes approach (GRASP-DR-RVND) is used.},
keywords = {Transportation, Internal combustion engines, Evolutionary computation, Electric vehicles, Routing, Batteries, Random variables, electric vehicle, constructive algorithm, location routing problem, battery swapping, green logistic, GRASP-RVND},
doi = {doi:10.1109/CEC45853.2021.9504799},
  notes = {Also known as \cite{9504799}},
)


@INPROCEEDINGS(Yang:2021:CEC,
%xplor 25 Aug 2021
  author = {Yifan Yang and Gang Chen and Hui Ma and Mengjie Zhang and Victoria Huang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Budget and {SLA} Aware Dynamic Workflow Scheduling in Cloud
Computing with Heterogeneous Resources},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2141--2148},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Workflow with different patterns and sizes arrive at a cloud
data center dynamically to be processed at virtual machines in the data
center, with the aim to minimize overall cost and makespan while
satisfying Service Level Agreement (SLA) requirement. To efficiently
schedule workflows, manually designed heuristics are proposed in the
literature. However, it is time consuming to manually design heuristics.
The designed heuristics may not work effectively for heterogeneous
workflow since only simple problem related factors are considered in the
heuristics. Further, most of the existing approaches ignore the deadline
constraints set in SLAs. Genetic Programming Hyper Heuristic (GPHH) can
be used to automatically design heuristics for scheduling problems. In
this paper, we propose a GPHH approach to automatically generate
heuristics for the dynamic workflow scheduling problem, with the goal of
minimizing the VM rental fees and SLA penalties. Experiments have been
conducted to evaluate the performance of the proposed approach. Compared
with several existing heuristics and conventional Genetic Programming
(GP) approaches, the proposed Dynamic Workflow Scheduling Genetic
Programming (DWSGP) has better performance and is highly adaptable to
variations in cloud environment.},
keywords = {Cloud computing, Schedules, Data centers, Processor scheduling, Heuristic algorithms, Computational modeling, Genetic programming, dynamic workflow scheduling, cloud computing, genetic programming},
doi = {doi:10.1109/CEC45853.2021.9504709},
  notes = {Also known as \cite{9504709}},
)


@INPROCEEDINGS(Li:2021:CEC,
%xplor 25 Aug 2021
  author = {Lianghao Li and Cheng He and Ran Cheng and Linqiang Pan},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Large-scale Multiobjective Optimization via Problem Decomposition
and Reformulation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2149--2155},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Large-scale multiobjective optimization problems (LSMOPs) are
challenging for existing approaches due to the complexity of objective
functions and the massive volume of decision space. Some large-scale
multiobjective evolutionary algorithms (LSMOEAs) have recently been
proposed, which have shown their effectiveness in solving some
benchmarks and real-world applications. They merely focus on handling
the massive volume of decision space and ignore the complexity of LSMOPs
in terms of objective functions. The complexity issue is also important
since the complexity grows along with the increment in the number of
decision variables. Our previous study proposed a framework to
accelerate evolutionary large-scale multiobjective optimization via
problem reformulation for handling large-scale decision variables. Here,
we investigate the effectiveness of LSMOF combined with
decomposition-based MOEA (MOEA/D), aiming to handle the complexity of
LSMOPs in both the decision and objective spaces. Specifically, MOEA/D
is embedded in LSMOF via two different strategies, and the proposed
algorithm is tested on various benchmark LSMOPs. Experimental results
indicate the encouraging performance improvement benefited from the
solution of the complexity issue in large-scale multiobjective
optimization.},
keywords = {Sociology, Evolutionary computation, Benchmark testing, Linear programming, Complexity theory, Statistics, Optimization, Multiobjective optimization, problem decomposition, large-scale optimization, offspring generation, problem reformulation},
doi = {doi:10.1109/CEC45853.2021.9504820},
  notes = {Also known as \cite{9504820}},
)


@INPROCEEDINGS(Zheng:2021:CEC,
%xplor 25 Aug 2021
  author = {Junhao Zheng and Lingjie Li and Qiuzhen Lin and Zhong Ming},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Improved Weighted Optimization-based Framework for Large-scale
MOPs},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2156--2163},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This paper proposes an improved weighted optimization-based
framework (iWOF) for solving large-scale multiobjective optimization
problems (LSMOPs). Compared to the original framework, there are two
main contributions in our work. Firstly, a novel evolutionary search
strategy involving two different search operators with different search
characteristics, i.e., a particle swarm optimization (PSO) operator and
differential evolution (DE) operator, is designed in iWOF, which aims to
provide a robust search ability on finding optimal solutions in a huge
decision space. Secondly, different from the stage in the original
framework that divides the whole evolutionary process into two
independent stages, the evolutionary process in the proposed iWOF is
simplified to only one stage, which effectively reduces the number of
predefined parameters. Besides that, the evolving numbers of weight
optimization and original optimization in iWOF are adjusted adaptively
according to the evolutionary stage. The experimental results on three
different groups of benchmark LSMOPs validate the superiority of the
proposed iWOF over WOF and other several state-of-the-art multiobjective
evolutionary algorithms.},
keywords = {Benchmark testing, Search problems, Particle swarm optimization, Optimization, Large-scale optimization problem (LSMOP), Multiobjective optimization algorithm, Particle swarm optimization (PSO), Differential evolution (DE)},
doi = {doi:10.1109/CEC45853.2021.9504854},
  notes = {Also known as \cite{9504854}},
)


@INPROCEEDINGS(Cao:2021:CEC,
%xplor 25 Aug 2021
  author = {Zijian Cao and Chen Liu and Zhenyu Wang and Haowen Jia},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Improved {MOEA/D} Framework with Multoperator Strategies for
Multi-objective Optimization Problems with a Large Scale of Variables},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2164--2170},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={A novel multi-objective evolutionary algorithm based on
decomposition is proposed, in which a multi-population meta-heuristic
algorithm named phase based optimization (PBO) is utilized as an
effective search engine for enhancing the performance of the original
MOEA/D, and is consequently termed MOEA/D-PBO. Three different
sub-populations are divided by using the distance between the
individuals and the reference point in accordance with the principle
from high to low. On this basis, three different search strategies are
executed to generate the better solutions according to the position
characteristic of individuals with the reference point as the center.
The empirical results demonstrate that MOEA/D-PBO can provide much
better performance on large scale bi-objective and tri-objective
optimization problems than MOEA/D-DE and IM-MOEA.},
keywords = {Evolutionary computation, Search engines, Search problems, Task analysis, Optimization, Decomposition, Multi-population, Large Scale Optimization, Multi-objective Optimization},
doi = {doi:10.1109/CEC45853.2021.9504768},
  notes = {Also known as \cite{9504768}},
)


@INPROCEEDINGS(Liu:2021:CEC,
%xplor 25 Aug 2021
  author = {Fei-Yu Liu and Chao Qian},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Prediction Guided Meta-Learning for Multi-Objective Reinforcement
Learning},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2171--2178},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Many real-world control problems consist of several different,
possibly conflicting, objectives, which require finding a high-quality
set of policies that are optimal for different objective preferences.
Extensive research mainly focused on how to obtain a high-quality
approximated Pareto set of policies, while another important research
direction studies how to adapt to new objective preferences quickly. In
this paper, we propose a new multi-objective reinforcement learning
(MORL) algorithm so-called PG-Meta-MORL for achieving both goals.
PG-Meta-MORL frames MORL as a meta-learning problem and iteratively
optimizes a meta-policy using multiple tasks with objective preferences
selected based on a prediction model, which is trained to guide the
optimization process towards best improving the quality of the current
Pareto set of policies. The empirical results on several multi-objective
continuous control problems show that PG-Meta-MORL can find a
high-quality approximated Pareto set of policies, and meanwhile, the
obtained meta-policy can be adapted well to new objective preferences
using few-shot interactions with the environment.},
keywords = {Adaptation models, Reinforcement learning, Evolutionary computation, Predictive models, Prediction algorithms, Approximation algorithms, Task analysis},
doi = {doi:10.1109/CEC45853.2021.9504972},
  notes = {Also known as \cite{9504972}},
)


@INPROCEEDINGS(Madani:2021:CEC,
%xplor 25 Aug 2021
  author = {Amirali Madani and Beatrice Ombuki-Berman and Andries Engelbrecht},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Decision Space Scalability Analysis of Multi-Objective Particle
Swarm Optimization Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2179--2186},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Particle swarm optimization (PSO) has been adapted to solve
multi-objective optimization problems. However, these PSO-based
multi-objective optimization algorithms typically face difficulties when
the number of decision variables is increased and the problems turn into
large-scale multi-objective problems (LSMOPs). This paper presents a
decision space scalability analysis of five PSO-based multi-objective
optimization algorithms, namely optimized multi-objective particle swarm
op-timization (OMOPSO), speed-constrained multi-objective particle swarm
optimization (SMPSO), multi-objective particle swarm optimization with
multiple search strategies (MMOPSO), multi-guide particle swarm
optimization (MGPSO), and competitive mechanism-based multi-objective
particle swarm optimization (CMOPSO) for 24, 50, 100, 500 and 1000
dimensions (decision variables) to see how well each one of the
algorithms scales as the number of decision variables is increased. The
results indicate that, with an increase in the number of decision
variables, MMOPSO and SMPSO had the best scalability, each dominating
specific functions. Moreover, despite MGPSO's competitive performance on
the 24-dimensional functions, it showed the worst overall scalability
together with CMOPSO.},
keywords = {Scalability, Search problems, Particle swarm optimization, Optimization, Faces, Particle Swarm Optimization (PSO), Multi-Objective Optimization, Multi-Objective Evolutionary Algorithms (MOEAs)},
doi = {doi:10.1109/CEC45853.2021.9504846},
  notes = {Also known as \cite{9504846}},
)


@INPROCEEDINGS(da-Silva:2021:CEC,
%xplor 25 Aug 2021
  author = {Cleber A.C.F. {da Silva} and Daniel {Carneiro Rosa} and Pericles B.C. Miranda and Filipe R. Cordeiro and Tapas Si and Andre C.A. Nascimento and Rafael F. L. Mello and Paulo S. G. {de Mattos Neto}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multi-Objective Grammatical Evolution Framework to Generate
Convolutional Neural Network Architectures},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2187--2194},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Deep Convolutional Neural Networks (CNNs) have reached the
attention in the last decade due to their successful application to many
computer vision domains. Several handcrafted architectures have been
proposed in the literature, with increasing depth and millions of
parameters. However, the optimal architecture size and parameters setup
are dataset-dependent and challenging to find. For addressing this
problem, this work proposes a Multi-Objective Grammatical Evolution
framework to automatically generate suitable CNN architectures (layers
and parameters) for a given classification problem. For this, a
Context-free Grammar is developed, representing the search space of
possible CNN architectures. The proposed method seeks to find suitable
network architectures considering two objectives: accuracy and F1-score.
We evaluated our method on CIFAR-10, and the results obtained show that
our method generates simpler CNN architectures and overcomes the results
achieved by larger (more complex) state-of-the-art CNN approaches and
other grammars.},
keywords = {Computer vision, Computer architecture, Evolutionary computation, Network architecture, Grammar, Convolutional neural networks, Optimization, Grammatical Evolution, Deep Neural Networks, Multi-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504822},
  notes = {Also known as \cite{9504822}},
)


@INPROCEEDINGS(Tenstad:2021:CEC,
%xplor 25 Aug 2021
  author = {Amund Tenstad and Pauline C. Haddow},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{DES-HyperNEAT:} Towards Multiple Substrate Deep {ANNs}},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2195--2202},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Neuroevolution (NE), inspired by the natural evolution of
biological brains, may be applied to the evolution of both weights and
the topology of Artificial Neural Networks (ANNs). DES-HyperNEAT,
proposed herein, further extends existing NE algorithms to enable
evolvable substrate topologies with non-fixed node positions - a
desirable feature for optimisation of deep ANNs. DES-HyperNEAT builds on
HyperNEAT and ES-HyperNEAT whilst adding beneficial features from NEAT
and MSS-HyperNEAT. The preliminary experiments in this work consider
potential variants of DES-HyperNEAT, evaluating such variants in terms
of performance and efficiency on the Iris, Wine and Retina datasets.
Further, DES-HyperNEAT, ES-HyperNEAT, HyperNEAT and NEAT are compared to
highlight whether the DES-HyperNEAT extension has merit in terms of
performance and efficiency.},
keywords = {Iris, Network topology, Evolution (biology), Reinforcement learning, Evolutionary computation, Artificial neural networks, Retina, DES-HyperNEAT, Neuroevolution, Deep ANNs, ES-HyperNEAT, HyperNEAT, NEAT},
doi = {doi:10.1109/CEC45853.2021.9504803},
  notes = {Also known as \cite{9504803}},
)


@INPROCEEDINGS(Zhao:2021:CEC,
%xplor 25 Aug 2021
  author = {Liang Zhao and Wei Fang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Efficient and Flexible Automatic Search Algorithm for
Convolution Network Architectures},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2203--2210},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={As many superior convolutional neural networks (CNNs) have
been proposed in recent years, CNNs have played an important role in
computer vision. However, manually-designing CNN architecture is
difficult since expertise is required. Therefore, several automatic
search algorithms have been proposed for neural architecture search,
which usually have considerable computational complexity and the search
space is limited. To address these problems, an efficient and flexible
CNN architecture search algorithm (EF-CNN) is proposed in this paper. In
EF-CNN, a flexible architecture search space is constructed by
considering the depth, width, and lightweight blocks. In order to
improve the reliability of the architecture while reducing the
computational time, a multi-objective fitness correction method is
proposed in EF-CNN based on the divided datasets, where the accuracy and
computational complexity of architecture are considered simultaneously
to design CNN. The experimental results on CIFAR-10 and CIFAR-100
indicate that the performance of CNN architecture designed by EF-CNN is
very competitive while the computational time is greatly reduced.},
keywords = {Computer vision, Convolution, Computer architecture, Manuals, Evolutionary computation, Network architecture, Search problems, Convolutional neural network, Neural Architectures Search, Multi-Objective},
doi = {doi:10.1109/CEC45853.2021.9504945},
  notes = {Also known as \cite{9504945}},
)


@INPROCEEDINGS(Sanjeewani:2021:CEC,
%xplor 25 Aug 2021
  author = {Pubudu Sanjeewani and Brijesh Verma and Joseph Affum},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Novel Evolving Classifier with a False Alarm Class for Speed
Limit Sign Recognition},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2211--2217},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Automatic interpretation of information written on roadside
signs is very useful in applications for conducting road surveys,
driving autonomous vehicles, and improving the road safety and
infrastructure. However, it is a very difficult problem because of high
similarity and poor visibility of signs and challenging nature of road
environment. In this paper, we propose a novel technique for automatic
detection and recognition of speed limit signs. The proposed technique
contains two novel concepts. Firstly, pixel-wise segmentation of all
speed limit signs into one class instead of each sign in a separate
class is proposed. Secondly, a novel classifier with a false alarm class
that evolves its weights so that it can distinguish speed signs from
non-speed signs is proposed. The proposed technique is evaluated on a
real-world dataset provided by our industry partners. The dataset has
been created from videos of all state roads in Queensland. The
comparative analysis of results showed that the proposed technique is
able to detect and classify speed limit signs with high accuracy.},
keywords = {Industries, Convolution, Neural networks, Evolutionary computation, Road safety, Convolutional neural networks, Optimization, Deep learning, speed limit sign recognition, fully convolutional neural network, genetic optimization of convolution neural networks},
doi = {doi:10.1109/CEC45853.2021.9504710},
  notes = {Also known as \cite{9504710}},
)


@INPROCEEDINGS(Baioletti:2021:CEC,
%xplor 25 Aug 2021
  author = {Marco Baioletti and Gabriele Di Bari and Valentina Poggioni and Carlos Artemio Coello Coello},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Smart Multi-Objective Evolutionary {GAN}},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2218--2225},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Generative Adversarial Network (GAN) is a family of machine
learning algorithms designed to train neural networks able to imitate
real data distributions. Unfortunately, GAN suffers from problems such
as gradient vanishing and mode collapse. In Multi-Objective Evolutionary
Generative Adversarial Network (MO-EGAN) these problems were addressed
using an evolutionary technique combined with Multi-Objective selection,
obtaining better results on synthetic datasets at the expense of larger
computation times. In this works, we present the Smart MultiObjective
Evolutionary Generative Adversarial Network (SMO-EGAN) algorithm, which
reduces the computational cost of MO-EGAN and achieves better results on
real data distributions.},
keywords = {Machine learning algorithms, Neural networks, Evolutionary computation, Generative adversarial networks, Computational efficiency},
doi = {doi:10.1109/CEC45853.2021.9504858},
  notes = {Also known as \cite{9504858}},
)


@INPROCEEDINGS(Rodriguez-de-la-Cruz:2021:CEC,
%xplor 25 Aug 2021
  author = {Juan-Antonio Rodriguez-de-la-Cruz and Hector-Gabriel Acosta-Mesa and Efren Mezura-Montes},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolution of Generative Adversarial Networks Using {PSO} for
Synthesis of COVID-19 Chest X-ray Images},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2226--2233},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The use of biomedical images for the training of various Deep
Learning (DL) systems oriented to health has reported a competitive
performance. However, DL needs a large number of images for a correct
generalization and, particularly in the case of biomedical images, these
can be scarce. Generative Adversarial Networks (GANs) as Data Augmenting
tools have reaped significant results to improve performance in tasks
that involve the use of this kind of image. However, the architectural
design of these generative models in the biomedical image area has been
usually relegated to the expertise of researchers. Moreover, GANs are
affected by training instability that may lead to poor quality results.
This paper presents a neuroevolution algorithm based on Particle Swarm
Optimization for the design and training of GANs for the generation of
biomedical Chest X-Ray (CXR) images of pneumonia caused by COVID-19. The
proposed approach allows having a swarm of GANs topologies, where each
one of them grows progressively while being trained at the same time.
The fitness value is based on the Frechet Inception Distance (FID). The
proposed algorithm is able to obtain better FID results compared to
handcrafted GANs for the synthesis of CXR images.},
keywords = {Training, COVID-19, Pulmonary diseases, Tools, Generative adversarial networks, Topology, Task analysis, GANs neuroevolution, Particle Swarm Optimization, Biomedical Imaging Synthesis, COVID-19},
doi = {doi:10.1109/CEC45853.2021.9504743},
  notes = {Also known as \cite{9504743}},
)


@INPROCEEDINGS(McBride:2021:CEC,
%xplor 25 Aug 2021
  author = {Ryan McBride and Herbert H. Tsang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{SIMARD-LinearFold:} Long Sequence {RNA} Design with Simulated
Annealing},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2234--2241},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={RNA Design is an essential problem in bioinformatics to tailor
RNA Sequences that guide our biology and medicine. While there are many
RNA Design solutions in literature, the field has been limited by the
high runtime to estimate how a candidate RNA sequence will fold,
especially with "long sequences" over 750 "bases"; as these folding
algorithms are called millions of times per RNA Design problem, this
limits RNA Design to shorter sequences despite the prevalence of long
sequences in nature; for example, computationally designing COVID's over
20000 bases RNA sequence is not feasible with current RNA Design
algorithms. To address this issue, we are the first work to integrate
LinearFold, a faster RNA prediction algorithm used by Baidu to analyze
COVID with a higher efficiency, into RNA Design. We compare the runtime
and solution quality of our Applied Research Lab's Simulated Annealing
solution (SIMARD) with and without LinearFold to design sequences
thousands of bases long that timeout after weeks of runtime with current
algorithms. We also survey challenges in terms of solution quality and
adjusting the cooling schedule parameters. This work is thus a first
step into RNA Design for longer RNA sequences.},
keywords = {COVID-19, Schedules, Runtime, Cooling, RNA, Web and internet services, Simulated annealing, Bioinformatics, RNA Design, Simulated Annealing, Evolutionary Algorithms, Computational Intelligence},
doi = {doi:10.1109/CEC45853.2021.9504978},
  notes = {Also known as \cite{9504978}},
)


@INPROCEEDINGS(Ahmad:2021:CEC,
%xplor 25 Aug 2021
  author = {Shadab Ahmad and Rajarshi Pal and Avatharam Ganivada},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Score Level Fusion of Multimodal Biometrics Using Genetic
Algorithm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2242--2250},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Multimodal biometric system fuses information from multiple
biometric modalities to overcome limitations of unimodal biometric
system. This fusion significantly enhances the performance of the
system. One of the ways of fusing information for multimodal biometrics
is score level fusion. In this paper, a novel score level fusion method
is proposed. Here, fusion at score level is formulated as an
optimization problem. The paper proposes a genetic algorithm (GA) based
approach to solve this optimization problem. It minimizes the distances
between an aggregated score list and each input score list from
individual biometric modality. The proposed GA based method uses
weighted Spearman footrule distance metric to compute the distance
between a pair of score lists. Superiority of the proposed method over
several state-of-the-art score level and rank level fusion methods is
demonstrated experimentally.},
keywords = {Measurement, Biometrics (access control), Fuses, NIST, Particle swarm optimization, Optimization, Biological cells, multimodal biometrics, score level fusion, genetic algorithm},
doi = {doi:10.1109/CEC45853.2021.9504927},
  notes = {Also known as \cite{9504927}},
)


@INPROCEEDINGS(Echtenbruck:2021:CEC,
%xplor 25 Aug 2021
  author = {Patrick Echtenbruck and Michael Emmerich and Martina Echtenbruck and Boris Naujoks},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Optimally Weighted Ensembles in Model-Based Regression for Drug
Discovery},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2251--2258},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In drug discovery, classification is a well established in
silico method based on machine learning algorithms. However, since the
activity value on a target protein is described as a continuous value, a
regressional approach is worth being considered as well. The results of
the regression can then be turned into classification results with a
given threshold value. To further improve the results, a new method
called optimally weighted ensembles that uses a combination of more than
one model to build a better performing ensemble of models, is applied
here. This method is used in the context of drug discovery for the first
time. Naturally, it is crucial to choose suitable models for the
specific problem, if any prior knowledge is available. Statistical
significance of the approach is verified using a second dataset with a
different target. In this work, we show to what extent the obtained
classification results compare to previous, highly optimized, single
model results as presented in previous work. Furthermore, the results
are compared to ensembles where none of the contributing models were
optimized beforehand. All case studies are performed using the publicly
available database ChEMBL^1 .},
keywords = {Drugs, Proteins, Machine learning algorithms, Databases, Computational modeling, Biological system modeling, Machine learning, Bioactivity Prediction, In-Silico Drug Discovery, ChEMBL, Evolutionary Algorithm, Optimal Ensembles, Optimization for Machine Learning},
doi = {doi:10.1109/CEC45853.2021.9504835},
  notes = {Also known as \cite{9504835}},
)


@INPROCEEDINGS(Witten:2021:CEC,
%xplor 25 Aug 2021
  author = {Matthew Witten and Owen Clancey},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={The Effect of Mask Usage on Viral Immune Escape Times: An
Evolutionary Strategies-Inspired Model},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2259--2264},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The COVID-19 pandemic has posed many challenges to the world
in several spaces, including the public health, economic, and political
spheres. At the start of the pandemic, non-pharmaceutical interventions
(NPIs) such as lockdowns, social distancing, and mask-wearing were the
only tools available to combat transmission of SARS-CoV-2, the virus
that causes COVID-19. More recently, several vaccines have been approved
for use, and mass inoculation of the population has begun in many
countries.One pressing issue is the continued use of NPIs in the context
of vaccine deployment: is there an advantage to perpetuating the use of
these interventions even as inoculation proceeds? The present study aims
to address this question, investigating the impact of masking on the
time to viral immune evasion, i.e. how masking affects the length of
time the immune response is effective against the virus before the virus
mutates such that it can escape the immune system. Viral immune escape
is formulated as an optimization problem. A simplified and idealized
model of virus transmission is proposed, and an evolutionary strategies
approach is used to model the evolution of the virus.Results indicate
that there is strong justification for the continued use of masks, even
as vaccines are deployed. Immune escape times for no masking, 5percent of the
population masking, and 10percent of the population masking are 17.15 ? 1.92
generations, 21.13 ? 2.33 generations, and 62.27 ? 13.68 generations
respectively. This increase in viral immune escape time attributable to
masking is most likely caused by decreased transmission, such that
significant increases in genetic diversity are required for viral immune
evasion.},
keywords = {COVID-19, Pandemics, Computational modeling, Sociology, Genomics, Coronaviruses, Vaccines, COVID-19, evolutionary strategies, nonpharmaceutical interventions},
doi = {doi:10.1109/CEC45853.2021.9504724},
  notes = {Also known as \cite{9504724}},
)


@INPROCEEDINGS(Kuo:2021:CEC,
%xplor 25 Aug 2021
  author = {Jason Kuo and John W. Sheppard},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Tournament Topology Particle Swarm Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2265--2272},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Particle swarm optimization (PSO) has become a popular
algorithm for performing global numerical optimization; however, it is
known that the topology of PSO has a large influence on its performance.
Topologies with high connectivity can have fast convergence, but they
are also susceptible to convergence to local minima. Topologies with low
connectivity may avoid converging to local minima and achieve high
quality solutions, but they tend to have slow convergence. In this
paper, we propose a novel PSO topology based on a single-elimination
tournament. In the proposed tournament topology, particles move up a
tree structure through a fitness-based tournament. PSO updates then
propagate information about the global best position from the top of the
tree to the bottom. Experimental results on eleven benchmark functions
show that the proposed topology can achieve both the high quality
solutions of low-connectivity topologies and the fast convergence of
high-connectivity topologies.},
keywords = {Benchmark testing, Topology, Particle swarm optimization, Optimization, Convergence},
doi = {doi:10.1109/CEC45853.2021.9504863},
  notes = {Also known as \cite{9504863}},
)


@INPROCEEDINGS(al-Rifaie:2021:CEC,
%xplor 25 Aug 2021
  author = {Mohammad Majid al-Rifaie},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Investigating Knowledge-Based Exploration-Exploitation Balance in
a Minimalist Swarm Optimiser},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2273--2280},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={One of the key challenges in evolutionary, swarm and
population-based optimisers is the balance between exploration and
exploitation. The reliance on both guided and stochastic search in these
algorithms allows researchers to take different approaches to the topic.
This work uses a minimalist, vector-stripped swarm optimiser to present
a theoretical analysis on the behaviour of the particles. Being a
population-based continuous optimiser, dispersive flies optimisation or
DEO, bears several similarities with the well-known particle swarm
optimisers, differential evolution algorithms and their bare-bones
variants. The distinctive feature of this algorithm is its sheer
reliance on particles positions to update the population. The minimalist
nature of the algorithm reduces the challenges of understanding
particles oscillation around constantly changing centres, particles'
influence on one another, and their trajectories. This work presents a
unified exploration-exploitation probability study which is derived from
six scenarios in order to examine the population's dimensional behaviour
in each iteration. This paves the way to propose and investigate
adaptable, diversity promoting mechanisms. The proposed methods, which
may be extendable to other optimisers, are then examined on a
comprehensive set of benchmarks, and finally applied to high-dimensional
tomographic reconstruction which is an important inverse problem in
medical and industrial imaging.},
keywords = {Knowledge based systems, Sociology, Tomography, Particle measurements, Trajectory, Statistics, Particle swarm optimization, Exploration, exploitation, diversity, dispersive flies optimisation, DFO},
doi = {doi:10.1109/CEC45853.2021.9504805},
  notes = {Also known as \cite{9504805}},
)


@INPROCEEDINGS(Ihara:2021:CEC,
%xplor 25 Aug 2021
  author = {Koya Ihara and Shohei Kato},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Novel Sampling Method with Levy Flight for Distribution-Based
Discrete Particle Swarm Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2281--2288},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We have proposed a novel sampling method (NS) for controlling
step sizes and incorporating Levy flight to distribution-based discrete
particle swarm optimizations (DDP-SOs), which are discrete extended
variants of particle swarm optimizations handling the continuous
parameters of probability distributions over the variable values instead
of directly handling discrete variables. Our previous work demonstrated
that NS improved all DDPSOs on function optimization problems. However,
on categorical problems, the NS did not improve DDPSOs designed for
integer problems. For more detailed investigations, we conducted
optimization experiments on NK landscapes. The results show that NS
improves the DDPSOs' efficiency and robustness to large solution space
and non-separable problems. Besides, we found that NS is effective on
integer DDPSOs even for categorical optimization in cases where decision
variables have a few states. In addition, the proposed methods were
tested on feature selection experiments and achieved superior results
compared to some evolutionary algorithms designed for feature selection.},
keywords = {Aerospace electronics, Feature extraction, Sampling methods, Minimization, Robustness, Probability distribution, Size control, particle swarm optimization, discrete optimization, function optimization, NK landscape, feature selection},
doi = {doi:10.1109/CEC45853.2021.9504684},
  notes = {Also known as \cite{9504684}},
)


@INPROCEEDINGS(Dennis:2021:CEC,
%xplor 25 Aug 2021
  author = {Cody Dennis and Beatrice M. Ombuki-Berman and Andries Engelbrecht},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Predicting Particle Swarm Optimization Control Parameters From
Fitness Landscape Characteristics},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2289--2298},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Selecting appropriate control parameters for the particle
swarm optimization algorithm can be extremely time consuming and
expensive, yet it is necessary in order to achieve optimal performance
on a problem. Despite its significance, the issue of control parameter
selection remains an open problem. This work leverages techniques from
the field of fitness landscape analysis to characterize a large suite of
benchmark problems. Extensive experimentation is performed to identify
strong control parameters for each problem, and machine learning
techniques are used to predict strong control parameters from the
characterization of a problem. The results demonstrate that good
generalization is possible with minimal training data. This suggests
that the cost of parameter selection can be significantly reduced.},
keywords = {Training, Analytical models, Adaptation models, Computational modeling, Neural networks, Training data, Benchmark testing, Particle swarm optimization, Control parameters, Fitness landscape analysis, Control parameter configuration, Neural networks},
doi = {doi:10.1109/CEC45853.2021.9505006},
  notes = {Also known as \cite{9505006}},
)


@INPROCEEDINGS(Harrison:2021:CEC,
%xplor 25 Aug 2021
  author = {Kyle Robert Harrison and Beatrice M. Ombuki-Berman and Andries P. Engelbrecht},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Visualizing and Characterizing the Parameter Configuration
Landscape of Particle Swarm Optimization using Physical Landform
Classification},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2299--2306},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={When designing effective parameter tuning and/or self-adaptive
mechanisms for meta-heuristic optimizers, any insights about the
configuration process and its associated landscape are of great benefit.
Recently, the parameter configuration landscape (PCL) was proposed as a
mechanism to formally study and characterize the landscape induced by
the control parameter values of meta-heuristic search techniques. As an
extension, the use of geomorphon landform types to further characterize
and visualize the PCL was recently proposed. This study adopts the
geomorphon classification scheme and applies it to particle swarm
optimization (PSO). The methodology is applied on 20 minimization
benchmark problems with various problem dimensions and swarm sizes,
thereby providing deep insights into the PCL associated with PSO.},
keywords = {Visualization, Benchmark testing, Minimization, Particle swarm optimization, Tuning, Parameter configuration landscape, particle swarm optimization, fitness landscape analysis, ruggedness},
doi = {doi:10.1109/CEC45853.2021.9504760},
  notes = {Also known as \cite{9504760}},
)


@INPROCEEDINGS(O'Neill:2021:CEC,
%xplor 25 Aug 2021
  author = {Michael O'Neill and Anthony Brabazon},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Exploration of Asocial and Social Learning in the Evolution of
Variable-length Structures},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2307--2314},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={We wish to explore the contribution that asocial and social
learning might play as a mechanism for self-adaptation in the search for
variable-length structures by an evolutionary algorithm. An extremely
challenging, yet simple to understand problem landscape is adopted where
the probability of randomly finding a solution is approximately one in a
trillion. A number of learning mechanisms operating on variable-length
structures are implemented and their performance analysed. The social
learning setup, which combines forms of both social and asocial learning
in combination with evolution is found to be most performant, while the
setups exclusively adopting evolution are incapable of finding solutions.},
keywords = {Learning systems, Genomics, Genetic programming, Evolutionary computation, Search problems, Approximation algorithms, Encoding},
doi = {doi:10.1109/CEC45853.2021.9504900},
  notes = {Also known as \cite{9504900}},
)


@INPROCEEDINGS(Mohamed:2021:CEC,
%xplor 25 Aug 2021
  author = {Reem E. Mohamed and Saber Elsayed and Robert Hunjet and Hussein Abbass},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Graph-based Approach for Shepherding Swarms with Limited
Sensing Range},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2315--2322},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Within applications of swarm control, failing to maintain
cohesion amongst the agents may lead to mission failure. We study the
effect of limited sensing range of swarming agents when guided by a
shepherd. A connectivity-aware approach is proposed to enhance the
cohesion of the swarm. We combine a graph-based model of the flock with
particle swarm optimization (assisted by DBSCAN) to improve the
shepherd's performance in herding sheep (swarm members). The approach is
evaluated using multiple initial swarm configurations. Simulation
results on swarm sizes of 50 and 100 agents show up to 5percent reduction in
the task completion time and an average improvement of 2percent in the
success rate for low density sheep initialization scenarios and
competitive results for the higher density scenarios.},
keywords = {Simulation, Sensors, Task analysis, Particle swarm optimization, Shepherding, sensing range, particle swarm optimization, swarm control, sensing induced graph, DBSCAN},
doi = {doi:10.1109/CEC45853.2021.9504706},
  notes = {Also known as \cite{9504706}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {Stephen Chen and Shehnaz Islam and Antonio Bolufe-Rohler and James Montgomery and Tim Hendtlass},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Random Walk Analysis of Search in Metaheuristics},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2323--2330},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Random walks are a useful modeling tool for stochastic
processes. The addition of model features (e.g. finite travel in one
direction) can provide insight into specific practical situations (e.g.
gambler's ruin). A series of random walk experiments are designed to
study the effects of selection, exploration, and exploitation during the
search processes of metaheuristics. We present a set of random walk
conditions which leads to greater movement as the dimensionality of the
sampling distributions increases. We then implement a version of
Simulated Annealing in a similar search space which also achieves
improving performance with increasing dimensionality. Conversely, we
show that standard Particle Swarm Optimization has decreasing
performance with increasing dimensionality which is consistent with the
expected effects of the Curse of Dimensionality. These experiments give
us insights into future methods that metaheuristics might be able to
employ to defeat the Curse of Dimensionality (in globally convex,
continuous domain search spaces).},
keywords = {Stochastic processes, Simulated annealing, Tools, Particle swarm optimization, Standards, random walk, exploration, selection, metaheuristic, curse of dimensionality},
doi = {doi:10.1109/CEC45853.2021.9504687},
  notes = {Also known as \cite{9504687}},
)


@INPROCEEDINGS(Chaves:2021:CEC,
%xplor 25 Aug 2021
  author = {Antonio Augusto Chaves and Luiz Henrique Nogueira Lorena},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Adaptive and Near Parameter-Free {BRKGA} Using Q-Learning Method},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2331--2338},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Biased Random-Key Genetic Algorithm (BRKGA) is an
efficient metaheuristic to solve combinatorial optimization problems but
requires parameter tuning so the intensification and diversification of
the algorithm work in a balanced way. There is, however, not only one
optimal parameter configuration, and the best configuration may differ
according to the stages of the evolutionary process. Hence, in this
research paper, a BRKGA with Q-Learning algorithm (BRKGA-QL) is
proposed. The aim is to control the algorithm parameters during the
evolutionary process using Reinforcement Learning, indicating the best
configuration at each stage. In the experiments, BRKGA-QL was applied to
the symmetric Traveling Salesman Problem, and the results show the
efficiency and competitiveness of the proposed algorithm.},
keywords = {Process control, Reinforcement learning, Evolutionary computation, Traveling salesman problems, Tuning, Genetic algorithms, Genetic Algorithm, Q-Learning, Reinforcement Learning, Parameter control},
doi = {doi:10.1109/CEC45853.2021.9504766},
  notes = {Also known as \cite{9504766}},
)


@INPROCEEDINGS(Rodriguez-Zbinden:2021:CEC,
%xplor 25 Aug 2021
  author = {Sebastian Rodriguez-Zbinden and Elizabeth Montero and Carola Blazquez and Pablo Miranda},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Multi-objective Iterated Local Search Approach to Solve the
Insular Traveling Salesman Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2339--2346},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In this work, we solve the bi-objective insular traveling
salesman problem, which arises when a barge collects the waste generated
by a set of islands. It considers selection and routing decisions coming
from the selection of docks to be visited and the sequence of visits
performed by the barge. Moreover, the cost of the tour performed by the
barge and the waste ground transportation costs are minimized in a
multi-objective approach. In this work, we propose a local search
multi-objective approach to find approximations of the Pareto sets of
efficient solutions. It uses an iterated local search method to find
solutions in several areas of the front based on a weighted sum
approach. We evaluated the performance of our proposal on a set of
real-world problem instances from twenty-one islands in the south of
Chile. Our results demonstrated the ability of the proposed approach to
find high-quality approximations of the Pareto sets for all the problem
instances evaluated in reduced times compared to an exact approach.},
keywords = {Land transportation, Boats, Evolutionary computation, Traveling salesman problems, Search problems, Routing, Proposals, multi-objective insular traveling salesman problem, local search, iterated local search},
doi = {doi:10.1109/CEC45853.2021.9504830},
  notes = {Also known as \cite{9504830}},
)


@INPROCEEDINGS(Pires:2021:CEC,
%xplor 25 Aug 2021
  author = {Otto Menegasso Pires and Rafael {de Santiago} and Jerusa Marchi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Two Stage Quantum Optimization for the School Timetabling Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2347--2353},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Timetabling is an NP-Hard problem that searches for periodic
scheduling of events that must meet a set of hard and soft constraints.
Because of the difficulty in finding an exact solution, the use of
heuristics to address the problem is a common practice. When only the
hard constraints are considered, the timetabling problem can be reduced
to graph vertex coloring. This similarity between both problems has
motivated the use of graph coloring heuristics as a means to solve the
timetabling problem. We propose the use of the Quantum Approximate
Optimization Algorithm as a heuristic to solve the school timetabling
problem. The QAOA is a hybrid quantum-classical algorithm that can be
used to address combinatorial optimization problems. We simulated QAOA
in a minimal example with 42 qubits using the Ket Quantum Programming
Language and our results showed that it is possible to apply QAOA to the
school timetabling problem.},
keywords = {Limiting, NP-hard problem, Heuristic algorithms, Computational modeling, Qubit, Evolutionary computation, Approximation algorithms, qaoa, quantum computing, school timetabling, timetabling problem},
doi = {doi:10.1109/CEC45853.2021.9504701},
  notes = {Also known as \cite{9504701}},
)


@INPROCEEDINGS(Vieira:2021:CEC,
%xplor 25 Aug 2021
  author = {Carlos Vieira and Adelson {de Araujo} and Jose E. Andrade and Leonardo C. T. Bezerra},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={iSklearn: Automated Machine Learning with irace},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2354--2361},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Automated algorithm engineering has become an important asset
for academia and industry. irace, for instance, is an algorithm
configurator (AC) that has successfully designed effective algorithms
for optimization problems. The major advantage of irace is combining
learning and parallelization, but no fully-functional automated machine
learning (AutoML) system powered by irace has yet been proposed. This is
rather striking, as some of the most relevant existing AutoML tools are
powered by ACs, of which irace is one of the most effective examples.In
this work, we propose iSklearn, an irace-powered AutoML system. Our
proposal improves existing work applying an AC to engineer a machine
learning (ML) pipeline. First, our configuration space represents a
minimalist pipeline template, demonstrating that simpler pipelines can
be competitive with elaborate approaches (e.g. ensembles). Second, our
configuration setup improves the application of AC-based AutoML to time
series (TS) problems, and is more flexible to fit other applications.We
evaluate iSklearn on three major ML domains, namely computer vision
(CV), natural language processing (NLP), and TS. Results prove
competitive to AUTOSKLEARN, a state-of-the-art AutoML system also built
on scikit-learn. Furthermore, the compositions of the pipelines devised
vary with the problem domain and dataset considered, providing further
evidence for the need of AutoML tools. We conclude our investigation
ablating through the proposed configuration space and setup to
understand their impact on the performance of iSklearn.},
keywords = {Training, Industries, Computer vision, Machine learning algorithms, Pipelines, Time series analysis, Tools},
doi = {doi:10.1109/CEC45853.2021.9504696},
  notes = {Also known as \cite{9504696}},
)


@INPROCEEDINGS(da-Costa:2021:CEC,
%xplor 25 Aug 2021
  author = {Calvin Rodrigues {da Costa} and Maria C. V. Nascimento},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Intelligent-Guided Adaptive Search For The Traveling Backpacker
Problem},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2362--2369},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The solution of combinatorial problems has been largely
performed by heuristics for their ability to obtain good solutions
faster than exact methods. In this paper, we propose heuristic methods
to approach a hard-to-solve combinatorial optimization problem. The
target routing problem is the recently proposed Traveling Backpacker
Problem (TBP), which has not been investigated by a heuristic method
yet. The difficulty in constructing feasible solutions for such a
problem drove us to approach the TBP by a metaheuristic with a learning
stage in the search, the Intelligent Greedy Adaptive Search (IGAS). To
validate the introduced solution methods, they were compared to the
exact solution obtained by CPLEX. Besides, we analyze isolate parts of
the methods to infer about the performance of the construction of the
solution and the local search, the main phases of IGAS. The results of
computational experiments show that IGAS outperformed the other
introduced methods in small-sized and medium-sized instances, being
competitive in larger instances. Moreover, IGAS presented reasonable
gaps to the solutions obtained by the exact commercial solver.},
keywords = {Adaptive systems, Neural networks, Evolutionary computation, Search problems, Routing, Traveling Backpacker Problem, GRASP, IGAS, neural network, Growing Neural Gas},
doi = {doi:10.1109/CEC45853.2021.9504937},
  notes = {Also known as \cite{9504937}},
)


@INPROCEEDINGS(Vilar-Dias:2021:CEC,
%xplor 25 Aug 2021
  author = {Joao L. Vilar-Dias and M. A. S. Galindo and Fernando B. Lima-Neto},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Cultural Weight-Based Fish School Search: A Flexible Optimization
Algorithm For Engineering},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2370--2376},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Many real-life engineering applications are optimization
problems. To find the best configuration of variables to minimize costs
and maximize efficiency are typically used engineering software as CAD,
CAE and CAM. In this context, Machine Learning can be used to automate
and improve this type of application. This despite, those searches not
seldomly evoke unreliable areas and suggest risky solutions. Because of
inaccuracies, volatility, unfeasibility, and specificities of real
environments, the easy incorporation of cultural practices (i.e.
normative, situational, domain and historic knowledge) and as well as
the production of multiple acceptable solutions for a problem is always
welcome, especially in Engineering. The present article put forward a
hybridization of a multi-modal algorithm (Weight-Based Fish School
Search - wFSS) with Cultural Algorithms' belief space. New cwFSS is able
to guide the optimization also considering normative knowledge from
experts, technical literature and problem domain readily available
knowledge to prevent the incorporation of constraints into the fitness
function. We also evaluated the use of temporal knowledge to guide the
simulation. The proposed method was tested in a thermal power plant
efficiency optimization and compared with standard wFSS and the Niching
Migratory Multi-Swarm Optimizer (NMMSO), winner of CEC'2015 niching
competition. As results, cwFSS has outperformed at times NMMSO about
time, fitness and variability, as well as traditional wFSS about time,
stability, safeness and variability of the multimodal solutions. By
avoiding penalties, the appropriation of a priori search directly into
the search can effectively and elegantly help better support for
engineering decisions.},
keywords = {Knowledge engineering, Solid modeling, Search problems, Fish, Stability analysis, Cultural differences, Optimization, Swarm Intelligence, Multimodal Optimization, Weight-based FSS, Cultural Algorithms, Optimization in Engineering},
doi = {doi:10.1109/CEC45853.2021.9504779},
  notes = {Also known as \cite{9504779}},
)


@INPROCEEDINGS(Roberto-Farah:2021:CEC,
%xplor 25 Aug 2021
  author = {Paulo {Roberto Farah} and Thaina Mariani and Enrique A. {da Roza} and Rogerio C. Silva and Silvia {Regina Vergilio}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Unsupervised Learning For Refactoring Pattern Detection},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2377--2384},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Software refactoring changes the structure of a program
without modifying its external behavior, generally intending to improve
software quality attributes. However, refactoring is a complex activity
and, many times, a composition of refactorings is necessary. Besides,
some code elements are refactored similarly, considering the kind and
frequency of refactorings applied. Works in the refactoring literature
usually investigate the impact and understanding of an individual
refactoring, neglecting that developers have to apply more than one
refactoring operation to reach their goals. There is a lack of studies
to identify and characterize refactoring patterns. To fulfill this gap,
this work explores the use of unsupervised learning, particularly
cluster analysis, to group elements (Java classes) that are refactored
similarly in software repositories. We used a total of 1435 projects and
applied the K-Means algorithm to group classes that received the same
refactoring with the same frequency. We obtained a set of seven
clusters. Then, the main refactoring compositions associated with each
cluster are analyzed to identify the corresponding pattern. Each pattern
is described and also characterized using a set of metrics. The great
majority of refactoring compositions include only one kind of
refactoring, applied with low frequency. If we consider compositions
including more than one type of refactorings, combinations of Extract
Superclass and Pull Up Method are the most frequent.},
keywords = {Measurement, Java, Software algorithms, Clustering algorithms, Software quality, Evolutionary computation, Unsupervised learning, Software refactoring, clustering algorithms, Mining Software Repositories},
doi = {doi:10.1109/CEC45853.2021.9504804},
  notes = {Also known as \cite{9504804}},
)


@INPROCEEDINGS(Moghadam:2021:CEC,
%xplor 25 Aug 2021
  author = {Mahshid Helali Moghadam and Golrokh Hamidi and Markus Borg and Mehrdad Saadatmand and Markus Bohlin and Bjorn Lisper and Pasqualina Potena},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Performance Testing Using a Smart Reinforcement Learning-Driven
Test Agent},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2385--2394},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Performance testing with the aim of generating an efficient
and effective workload to identify performance issues is challenging.
Many of the automated approaches mainly rely on analyzing system models,
source code, or extracting the usage pattern of the system during the
execution. However, such information and artifacts are not always
available. Moreover, all the transactions within a generated workload do
not impact the performance of the system the same way, a finely tuned
workload could accomplish the test objective in an efficient way.
Model-free reinforcement learning is widely used for finding the optimal
behavior to accomplish an objective in many decision-making problems
without relying on a model of the system. This paper proposes that if
the optimal policy (way) for generating test workload to meet a test
objective can be learned by a test agent, then efficient test automation
would be possible without relying on system models or source code. We
present a self-adaptive reinforcement learning-driven load testing
agent, RELOAD, that learns the optimal policy for test workload
generation and generates an effective workload efficiently to meet the
test objective. Once the agent learns the optimal policy, it can reuse
the learned policy in subsequent testing activities. Our experiments
show that the proposed intelligent load test agent can accomplish the
test objective with lower test cost compared to common load testing
procedures, and results in higher test efficiency.},
keywords = {Analytical models, Automation, Transfer learning, Decision making, Reinforcement learning, Knowledge representation, Evolutionary computation, performance testing, load testing, workload generation, reinforcement learning, autonomous testing},
doi = {doi:10.1109/CEC45853.2021.9504763},
  notes = {Also known as \cite{9504763}},
)


@INPROCEEDINGS(Ramirez:2021:CEC,
%xplor 25 Aug 2021
  author = {Aurora Ramirez and Pedro Delgado-Perez and Kevin J. Valle-Gomez and Inmaculada Medina-Bulo and Jose Raul Romero},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Interactivity in the Generation of Test Cases with Evolutionary
Computation},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2395--2402},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Test generation is a costly but necessary testing activity to
increase the quality of software projects. Automated testing tools based
on evolutionary computation principles constitute an appealing modern
approach to support testing tasks. However, these tools still find
difficulties to detect certain types of plausible faults in real-world
projects. Besides, recent studies have shown that, in general,
automatically-generated tests do not resemble those manually written
and, consequently, testers are reluctant to adopt them. We observe two
key issues, namely the opacity of the process and the lack of
cooperation with the tester, currently hampering the acceptance of
automated results. Based on these findings, we explore in this paper how
the interaction between current tools and expert testers would help
address the test case generation problem. More specifically, we identify
a number of interaction opportunities related to the object-oriented
test case design driven to boost their readability and detection power.
Using EvoSuite as base implementation, we present a proof of concept
focused on the possibility to integrate readability assessment of the
most promising test suites into a genetic algorithm.},
keywords = {Evolutionary computation, Manuals, Tools, Software, Proposals, Test pattern generators, Object recognition, software testing, interactive search-based software engineering, test generation, mutation testing},
doi = {doi:10.1109/CEC45853.2021.9504786},
  notes = {Also known as \cite{9504786}},
)


@INPROCEEDINGS(Benito-Parejo:2021:CEC,
%xplor 25 Aug 2021
  author = {Miguel Benito-Parejo and Mercedes G. Merayo},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Using Genetic Algorithms To Select Test Cases For Finite State
Machines With Timeouts},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2403--2410},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Testing software is an expensive and time consuming task. This
drawback increases in the case that the system presents timeouts which
affect the functional behaviour. In this case, it is necessary to invest
more time to apply the test cases. Thus, it is desirable to diminish the
amount of test cases to be applied and, consequently, the time devoted
to test the system, as long as the fault detection capacity of the test
cases is not affected. In this work, we introduce a genetic algorithm
that selects a subset of test cases from an initial test suite for
systems that present timeouts, with the goal of keeping a good level of
effectiveness. We report on several experiments performed to compare the
generated solutions with random selection and the original test suite.},
keywords = {Fault detection, Automata, Evolutionary computation, Software, Task analysis, Genetic algorithms, Testing, Genetic algorithms, Selection of test cases, Testing from FSMs, Timeouts},
doi = {doi:10.1109/CEC45853.2021.9504764},
  notes = {Also known as \cite{9504764}},
)


@INPROCEEDINGS(Ibias:2021:CEC,
%xplor 25 Aug 2021
  author = {Alfredo Ibias and Pablo Vazquez-Gomis and Miguel Benito-Parejo},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Coverage-Based Grammar-Guided Genetic Programming Generation of
Test Suites},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2411--2418},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Software testing is fundamental to ensure the reliability of
software. To properly test software, it is critical to generate test
suites with high fault finding ability. We propose a new method to
generate such test suites: a coverage-based grammar-guide genetic
programming algorithm. This evolutionary computation based method allows
us to generate test suites that conform with respect to a specification
of the system under test using the coverage of such test suites as a
guide. We considered scenarios for both black-box testing and white-box
testing, depending on the different criteria we work with at each
situation. Our experiments show that our proposed method outperforms
other baseline methods, both in performance and execution time.},
keywords = {Software testing, Software algorithms, Genetic programming, Evolutionary computation, Software, Software reliability, Genetic communication, Genetic Programming, Coverage, Software Testing},
doi = {doi:10.1109/CEC45853.2021.9504969},
  notes = {Also known as \cite{9504969}},
)


@INPROCEEDINGS(Chivilikhin:2021:CEC,
%xplor 25 Aug 2021
  author = {Daniil Chivilikhin},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Temporal Specification Mining for {IEC} 61499 Function Blocks Using
Evolutionary Algorithms and Model Checking},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2419--2426},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Model checking allows rigorously checking the compliance of a
software model with a given formal specification described with temporal
logics. In control applications, model checking is used for formal
verification of controller code, and temporal logic is used as a formal
specification language. However, not all industrial projects start with
the creation of a formal specification. On the contrary, often model
checking is introduced at later stages of project development, and
formal specifications need to be written for already existing legacy
code.In this work we developed a framework for automating the process of
formal specification preparation for IEC 61499 function blocks, a
programming language used in the area of industrial automation. The
developed framework is based on multiobjective evolutionary algorithms
and model checking, and allows inference of linear temporal logic
formulas for basic IEC 61499 function blocks. Our experiments show that
the temporal specifications generated by the proposed framework
sufficiently approximate existing, manually prepared specifications.},
keywords = {Computer languages, Automation, Evolutionary computation, Model checking, Software, Formal specifications, IEC Standards},
doi = {doi:10.1109/CEC45853.2021.9504765},
  notes = {Also known as \cite{9504765}},
)


@INPROCEEDINGS(Hao:2021:CEC,
%xplor 25 Aug 2021
  author = {Hao Hao and Aimin Zhou and Hu Zhang},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Approximated Domination Relationship based on Binary
Classifiers for Evolutionary Multiobjective Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2427--2434},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Preselection is an important strategy to improve evolutionary
algorithms' performance by filtering out unpromising solutions before
fitness evaluations. This paper introduces a pre-selection strategy
based on an approximated Pareto domination relationship for
multiobjective evolutionary optimization. For each objective, a binary
relation between each pair of solutions is constructed based on the
current population, and a binary classifier is built based on the binary
relation pairs. In this way, an approximated Pareto domination
relationship can be defined. When new trial solutions are generated, the
approximated Pareto domination is used to select promising solutions,
which shall be evaluated by the real objective functions. The new
preselection is integrated into two algorithms. The experimental results
on two benchmark test suites suggest that the algorithms with
preselection outperform their original ones.},
keywords = {Filtering, Computational modeling, Sociology, Training data, Evolutionary computation, Approximation algorithms, Linear programming, evolutionary multiobjective optimization, approximated Pareto domination, surrogate model, classification},
doi = {doi:10.1109/CEC45853.2021.9504781},
  notes = {Also known as \cite{9504781}},
)


@INPROCEEDINGS(Luo:2021:CEC,
%xplor 25 Aug 2021
  author = {Yongkang Luo and Kening Zhang and Haipeng Yang and Feng Liu and Shuai Luo and Lei Zhang and Xiaoyan Sun},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Reduced Mixed Representation Based Multi-Objective Evolutionary
Algorithm for Large-Scale Overlapping Community Detection},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2435--2442},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In recent years, the application of multi-objective
evolutionary algorithms (MOEAs) to overlapping community detection in
complex networks has been a hot research topic. However, the existing
MOEAs for detecting overlapping communities show poor scalability to
large-scale networks due to the fact that the encoding length of
individuals is usually equal to the number of all nodes in the network.
To this end, we suggest a reduced mixed representation based
multi-objective evolutionary algorithm named RMR-MOEA for large-scale
overlapping community detection, where the length of the individual is
recursively reduced as the evolution proceeds. Specifically, a mixed
representation is adopted for fast encoding and decoding the individual
in the population, which consists of two parts: one represents all
potential overlapping nodes and the other represents all non-overlapping
nodes. Then, in each individual length reduction, two strategies are
suggested to respectively shorten the length of each part in the mixed
representation, with the aim to greatly reduce the search space.
Finally, the experimental results on 10 real-world complex networks
demonstrate the effectiveness of the proposed RMR-MOEA in terms of both
detection performance and running time, especially on large-scale
networks.},
keywords = {Scalability, Sociology, Evolutionary computation, Complex networks, Encoding, Decoding, Indexes, Index Terms--large-scale complex network, multi-objective optimization, evolutionary algorithm, overlapping community detection, mixed representation},
doi = {doi:10.1109/CEC45853.2021.9504894},
  notes = {Also known as \cite{9504894}},
)


@INPROCEEDINGS(Showalter:2021:CEC,
%xplor 25 Aug 2021
  author = {Ian Showalter and Howard Schwartz and Sidney N. Givigi},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Evolutionary Inherited Neuromodulated Neurocontrollers with
Objective Weighted Ranking},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2443--2450},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={In the physical world, individuals compete against others
within their own population or separately evolving populations. Robotic
agents will soon face this coevolutionary and adversarial reality. Many
challenges not encountered in the evolution of single populations are
encountered in adversarial coevolution. These challenges can render
single evolutionary approaches either less effective or ineffective.
Most problems have a fundamental goal, but also feature secondary
desirable objectives. Here, evader agents in the pursuit-evasion game
that do not elude capture are effectively useless. Secondly, when
applied to evolve real robots online in a coevolutionary context,
non-optimal robots can be erratic and cause damage. Objective hierarchy
can be used to define the importance of each objective, and promote
quicker optimization of primary objectives such as evasion. The
Evolutionary Inherited Neuromodulated Neurocontroller (EINN) method
incorporates objective weighted ranking (OWR), a novel objective
hierarchy method that promotes optimization of the primary objective in
simultaneous multi-objective optimization. EINN is compared to the
previously demonstrated Lamarckian-inherited Neuromodulated
MultiObjective Evolutionary Neurocontroller (LNMOEN), and shown to be
effective in a single evolutionary context.},
keywords = {Neurocontrollers, Evolution (biology), Sociology, Games, Learning (artificial intelligence), Topology, Statistics, Artificial Neural Network, Autonomous Vehicle, Hebbian Learning, Multi-Objective, Neuroevolution, Neuromodulation, Objective Hierarchy, Objective Weighted Ranking, Pursuit-Evasion, Unsupervised Learning},
doi = {doi:10.1109/CEC45853.2021.9504892},
  notes = {Also known as \cite{9504892}},
)


@INPROCEEDINGS(Mishra:2021:CEC,
%xplor 25 Aug 2021
  author = {Sumit Mishra and Srinibas Swain and Sangita Sarmah and Carlos A. {Coello Coello}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Hypervolume by Slicing Objective Algorithm: An Improved Version},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2451--2458},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The hypervolume remains a popular performance indicator in
evolutionary multi-objective, mainly because of its nice mathematical
properties (i.e., it's the only performance indicator known to be
Pareto-compliant). However, its high computational cost (which grows
polynomially on the population size but exponentially on the number of
objectives) has severely limited its use in many-objective optimization.
This has motivated a variety of proposals that attempt to overcome this
limitation. One of the most popular proposals currently available is the
so-called Hypervolume by Slicing Objectives (HSO) algorithm. Here, we
show that the worst-case time complexity of the HSO algorithm, as
obtained by its authors, is incorrect. Then, we provide an efficient
implementation of the HSO algorithm, which guarantees that unique slices
are generated to compute the hypervolume.},
keywords = {Sociology, Evolutionary computation, Computational efficiency, Complexity theory, Proposals, Time complexity, Statistics, evolutionary computation, hypervolume, non-dominated points, time complexity},
doi = {doi:10.1109/CEC45853.2021.9504725},
  notes = {Also known as \cite{9504725}},
)


@INPROCEEDINGS(Cristina-Valencia-Rodriguez:2021:CEC,
%xplor 25 Aug 2021
  author = {Diana {Cristina Valencia-Rodriguez} and Carlos A. {Coello Coello}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Ensemble of Scalarizing Functions and Weight Vectors for
Evolutionary Multi-Objective Optimization},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2459--2467},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Ensembles have been used in the evolutionary computation
literature to evolve several populations in an independent manner, using
different search approaches. Moreover, each population's parents compete
with their offspring and the other population's offspring to improve
diversity. It has been shown that ensemble algorithms improve the
performance of the techniques embedded within them, when considered
independently. Furthermore, scalarizing functions have been successfully
used in decomposition-based and some indicator-based Multi-objective
Evolutionary Algorithms (MOEAs). However, it has been shown that the
performance of scalarizing function tends to be tied to the geometrical
shape of the Pareto front. In this work, we propose a new ensemble
algorithm that adopts different scalarizing functions and weight vectors
using Hungarian Differential Evolution as the baseline multi-objective
optimizer. Our experimental study shows that our proposed approach
outperforms the original HDE, and it is competitive with respect to
modern MOEAs.},
keywords = {Shape, Sociology, Evolutionary computation, Superluminescent diodes, Generators, Indexes, Statistics, Index Terms, ensemble, scalarizing functions, weight vectors, multi-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504941},
  notes = {Also known as \cite{9504941}},
)


@INPROCEEDINGS(Nan:2021:CEC,
%xplor 25 Aug 2021
  author = {Yang Nan and Ke Shang and Hisao Ishibuchi and Linjun He},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={A Two-stage Hypervolume Contribution Approximation Method Based
on R2 Indicator},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2468--2475},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Hypervolume-based multi-objective evolutionary algorithms
(HV-MOEAs) are one of the popular algorithm classes in the evolutionary
multi-objective optimization (EMO) com-munity. HV-MOEAs, which can
directly optimize the HV of a solution set, are useful in various
applications. However, the computation time of HV-MOEAs is very long for
many-objective problems since the calculation of the hypervolume
contribution (HVC) is computationally expensive. Therefore, a number of
approximation methods for the HVC calculation were proposed to reduce
its time cost. An R2-based hypervolume contribution approximation
(R2-HVC) method was proposed for HVC approximation. However, for
HV-MOEAs, the point is to find the worst solution, instead of accurately
approximating the HVC of each solution. In this paper, a novel method
(i.e., two-stage R2-HVC) is proposed for improving the ability of R2-HVC
to correctly identify the worst solution (i.e., the solution with the
smallest HVC value) in a solution set. In the proposed method, some
candidate solutions are selected based on rough HVC approximation in the
first stage, and they are carefully evaluated in the second stage. It is
shown through computational experiments that the proposed method
performs much better than the original R2-HVC method.},
keywords = {Evolutionary computation, Computational efficiency, Approximation methods, Indexes, Optimization, Guidelines, Index Terms, Multi-objective optimization, indicator-based algorithms, hypervolume, hypervolume contribution},
doi = {doi:10.1109/CEC45853.2021.9504726},
  notes = {Also known as \cite{9504726}},
)


@INPROCEEDINGS(de-Lima-Mendes:2021:CEC,
%xplor 25 Aug 2021
  author = {Raphael {de Lima Mendes} and Alexandre {Henrick da Silva Alves} and Matheus {de Souza Gomes} and Pedro {Luiz Lima Bertarini} and Laurence {Rodrigues do Amaral}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Many Layer Transfer Learning Genetic Algorithm ({MLTLGA):} a New
Evolutionary Transfer Learning Approach Applied To Pneumonia
Classification},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2476--2482},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Convolutional Neural Networks (CNNs) are considered the
state-of-the-art in computer vision tasks, being able to surpass human
performance in certain applications. The use of this technology in
medical applications for automated or aided diagnosis of diseases is
promising, especially in radiotherapy- based images. One of the most
deadly diseases, Pneumonia, can be diagnosed from chest x-ray images.
However, the use of Deep Learning relies on a huge amount of data to
build models. In order to leverage the transferable knowledge from
neural networks, transfer learning is usually used to create models from
pre-trained networks on different source datasets. In this paper, a
novel genetic algorithm (GA), Many Layer Transfer Learning Genetic
Algorithm (MLTLGA), is proposed to select the trainable layers of the
InceptionV3 CNN architecture to build a classifier of Pneumonia images
from chest x-ray images. The MLTLGA shows better performance than other
Genetic Algorithms in the literature and conventional transfer learning
methods, providing a promising framework for applications that rely on
Transfer Learning. The experiments show that the proposed method is at
least percent more accurate than similar evolutionary approaches and transfer
learning, more specifically, last layer training.},
keywords = {Training, Pulmonary diseases, Transfer learning, Neural networks, Medical services, Convolutional neural networks, Task analysis, Index Terms, Transfer Learning, Machine Learning, Evolutionary Computation, Convolutional Neural Networks, Genetic Algorithms, Artificial Intelligence},
doi = {doi:10.1109/CEC45853.2021.9504912},
  notes = {Also known as \cite{9504912}},
)


@INPROCEEDINGS(Mason:2021:CEC,
%xplor 25 Aug 2021
  author = {Karl Mason and Santiago Grijalva},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Building {HVAC} Control via Neural Networks and Natural Evolution
Strategies},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2483--2490},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Buildings are the highest consumers of electrical energy
globally. Developing effective control strategies to enhance the
operation of buildings to increase their energy efficiency has therefore
become a very active area of research. This research proposes the use of
evolutionary neural networks for the task of Heating Ventilation and Air
Conditioning control. A neural network controller is trained using the
exponential Natural Evolutionary Strategy algorithm is proposed. Its
performance is bench-marked against multiple state of the art control
strategies, including: a Genetic Algorithm trained neural network, the
popular Deep Q-Network reinforcement learning algorithm and also a
threshold based control policy. The results presented demonstrate that
evolutionary neural networks are an effective strategy for reducing
energy consumption while also minimizing thermal discomfort. The energy
consumption and thermal discomfort achieved by evolutionary neural
networks, are significantly lower than when using Deep Q-Network. This
is particularly true for the Natural Evolutionary Strategy algorithm.},
keywords = {Training, Energy consumption, HVAC, Neural networks, Buildings, Reinforcement learning, Evolutionary computation, HVAC, Buildings, Energy, Neural Networks, Natural Evolution Strategies, Reinforcement Learning, Evolutionary Algorithms},
doi = {doi:10.1109/CEC45853.2021.9504800},
  notes = {Also known as \cite{9504800}},
)


@INPROCEEDINGS(Frachon:2021:CEC,
%xplor 25 Aug 2021
  author = {Luc Frachon and Wei Pang and George M. Coghill},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Immune-Inspired Approach to Macro-Level Neural Ensemble Search},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2491--2498},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Recent years have seen a renewed interest in evolutionary
computation applied to the automatic design of deep neural network
architectures, i.e. Neural Architecture Search (NAS). The advantages of
evolutionary approaches in NAS include their conceptual simplicity and
their flexibility with regards to search space definition and/or
optimization objective.However, Artificial Immune Systems (AIS) that
follow the evolutionary computation paradigm are less explored in NAS.
In this research, we aim to leverage their intrinsic and excellent
ability to balance performance and population diversity to develop a
novel Neural Ensemble Search method, based on the Clonal Selection
Algorithm [1]. For more generality, we focus on designing
macro-architectures rather than architectural components.Experiments on
popular computer vision benchmarks demonstrate that our method reaches
competitive accuracy and efficiency despite minimal augmentation and
post-processing. We show that the AIS brings tangible benefits,
including maintaining the diversity of solutions, a semantically
straightforward implementation, and high efficiency. Moreover, this AIS
can exhibit a "secondary response": when presented with a related but
more difficult task, the ensemble will perform competently with zero
modification to the architectures or the training protocol.},
keywords = {Training, Protocols, Sociology, Evolutionary computation, Computer architecture, Search problems, Prediction algorithms, Deep Neural Networks, Artificial Immune Systems, Neural Ensemble Search},
doi = {doi:10.1109/CEC45853.2021.9504955},
  notes = {Also known as \cite{9504955}},
)


@INPROCEEDINGS(Phillips:2021:CEC,
%xplor 25 Aug 2021
  author = {Antin Phillips and Mathys C. {Du Plessis}},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Adaptive Navigation Strategies in Evolutionary Robotics
Controllers with Location Perception},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2499--2506},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={This research proposed augmenting Evolutionary Robotics
controllers with their own robotic simulator. The motivation behind this
idea is that the controller could use the simulator to approximate the
effects of its actions. This would allow the controller to maintain an
approximation of the robot's location and orientation. It was
hypothesised that this information could enhance a controller's
navigation ability. Both regular and augmented controllers were tested
and compared on a task that required them to employ adaptive navigation
strategies to efficiently solve the task. Augmented controllers were
shown to perform better and possess superior behaviours in comparison to
regular controllers.},
keywords = {Evolutionary robotics, Navigation, Neural networks, Robot sensing systems, Sensors, Task analysis, Erbium, Evolutionary Robotics, Neural Networks, Simulator Neural Networks, Proprioception, Path Integration, Location Perception},
doi = {doi:10.1109/CEC45853.2021.9504759},
  notes = {Also known as \cite{9504759}},
)


@INPROCEEDINGS(Chen:2021:CEC,
%xplor 25 Aug 2021
  author = {Xuhao Eric Chen and Brian J. Ross},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Deep Neural Network Guided Evolution of L-System Trees},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2507--2514},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Lindenmayer systems (L-systems) are mathematical formalisms
used for generating recursive structures. They are particularly
effective for defining realistic tree and plant models. It takes
experience to use L-systems effectively, however, as the final rendered
results are often difficult to predict. This research explores the use
of genetic programming (GP) and deep learning towards the automatic
evolution of L-system expressions that render 2D tree designs. As done
before by other researchers, the L-system language is easily defined and
manipulated by the GP system. It is challenging, however, to determine a
fitness function to evaluate the suitability of evolved expressions. We
train a deep convolutional neural network (CNN) to recognize suitable
trees rendered in the style of the L-system language. Experiments
explore a number of deep CNN strategies. Results in some experiments are
very promising, as images conforming to specified styles of tree species
were often produced. We found that underspecifying or over-complicating
the training requirements can arise, and the results become
unsatisfactory in such cases. Our results also confirm that of other
researchers, in that deep learning can be fooled by evolutionary
algorithms, and the criteria for success learned by deep neural networks
might not conform with those of human users.},
keywords = {Deep learning, Training, Solid modeling, Three-dimensional displays, Architecture, Vegetation, Evolutionary computation, genetic programming, convolutional neural networks, L-systems},
doi = {doi:10.1109/CEC45853.2021.9504827},
  notes = {Also known as \cite{9504827}},
)


@INPROCEEDINGS(Kulkarni:2021:CEC,
%xplor 25 Aug 2021
  author = {Divya D. Kulkarni and Shivashankar B. Nair},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={An Immuno-Inspired Transfer Learning Paradigm},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2515--2522},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Transfer learning in Artificial Neural Networks (ANN) is
performed by transferring some or all the layers of neurons from within
an already learned source ANN, to a target ANN which is then made to
learn a new domain. Though such transfers are known to accelerate the
convergence of the target ANN, there is no hard and fast method to
ascertain which of the neurons or layers need to be transferred. While
most methods advocate the transfer of complete layers of neurons, this
paper describes an Immuno-inspired transfer learning paradigm that aids
in ascertaining the pertinent or Hot neurons which when transferred
facilitate faster convergence of a target ANN. The method uses the
concept of an Idiotypic Network to evolve the temperatures associated
with neurons within the various layers of the source ANN. Temperatures
of those neurons which contribute more to the learning, increase making
them hotter. This paper also presents results obtained from experiments
performed on dissimilar datasets using this method, which clearly
authenticate its efficacy.},
keywords = {Neurons, Transfer learning, Artificial neural networks, Evolutionary computation, Biological neural networks, Convergence, Artificial Immune System, Idiotypic Network, Transfer Learning, Artificial Neural Network, Convolutional Neural Network},
doi = {doi:10.1109/CEC45853.2021.9504697},
  notes = {Also known as \cite{9504697}},
)


@INPROCEEDINGS(Ikushima:2021:CEC,
%xplor 25 Aug 2021
  author = {Naoya Ikushima and Keiko Ono and Yuya Maeda and Erina Makihara and Yoshiko Hanada},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Differential Evolution Neural Network Optimization with
Individual Dependent Mechanism},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2523--2530},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={With the increase of scenes where Neural Networks are used as
a classifier, the expectation of the classifying accuracy for the
network has risen. To improve classifying accuracy, Differential
Evolution (DE) has been applied as an optimization method for Neural
Networks. Compared to other DE methods, Differential Evolution with an
Individual-Dependent Mechanism (IDE) takes in account of the differences
between the fitness value of individuals. As a result, IDE has better
results in terms of accuracy. Therefore in this paper, a Neural Network
optimizer using IDE is proposed for further Neural Network improvement.
Moreover, while most optimizers would use the loss function as the
fitness value in DE, a method using accuracy is proposed because DE does
not require the activation function to be differentiable. Experiments
using the proposed framework has been conducted on classification
problems, and comparative studies with other self-adaptive mutations and
traditional optimizing methods have been performed. Experimental results
showed that the proposed method outperformed other conventional methods
in terms of accuracy.},
keywords = {Neural networks, Optimization methods, Evolutionary computation},
doi = {doi:10.1109/CEC45853.2021.9504713},
  notes = {Also known as \cite{9504713}},
)


@INPROCEEDINGS(Garcia-Valdez:2021:CEC,
%xplor 25 Aug 2021
  author = {Mario Garcia-Valdez and Rene Marquez and Leonardo Trujillo and J.J. Merelo},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Random Selection of Parameters in Asynchronous Pool-Based
Evolutionary Algorithms},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2531--2538},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Synchronous operation is not the most natural, as in
biologically inspired, mode to run distributed algorithms. In many grid,
cloud or volunteer setups nodes are heterogeneous, or simply are not
available at the exact same time; this is a challenge for the researcher
if their full performance is going to be actually leveraged.
Asynchronous distributed evolutionary algorithms try to solve this by
dropping the homogeneity, as well as the synchronicity, assumption.
These algorithms share the population between distributed workers which
execute the actual evolutionary process by taking samples of the
population, and replacing them in the population pool by evolved
individuals. The performance of these EAs depends in part on the
selection of parameters for the EA running in each worker. In this paper
we study how randomly varying parameters in distributed evolutionary
algorithms affects performance. Experiments were conducted in the AWS
cloud using 2, 6 and 12 virtual machine configurations, with both
homogeneous and heterogeneous random settings using five test functions
for real-valued optimization and the OneMax binary problem. The results
suggest that this method can produce a performance that is competitive
with instances of the algorithm using workers with parameters specially
tuned for the benchmark.},
keywords = {Cloud computing, Sociology, Evolutionary computation, Benchmark testing, Virtual machining, Biology, Statistics, Distributed Evolutionary Algorithms, Volunteer Computing, Cloud Computing, pool-based evolutionary algorithms, parameter optimization},
doi = {doi:10.1109/CEC45853.2021.9504880},
  notes = {Also known as \cite{9504880}},
)


@INPROCEEDINGS(Bouter:2021:CEC,
%xplor 25 Aug 2021
  author = {Anton Bouter and Tanja Alderliesten and Peter A.N. Bosman},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={{GPU-Accelerated} Parallel Gene-pool Optimal Mixing Applied to
Multi-Objective Deformable Image Registration},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2539--2548},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The Real-Valued Gene-pool Optimal Mixing Evolutionary
Algorithm (RV-GOMEA) has previously been successfully used to achieve
highly scalable optimization of various real-world problems in a
gray-box optimization setting. Deformable Image Registration (DIR) is a
multi-objective problem, aimed at finding the most likely non-rigid
deformation of a given source image so that it matches a given target
image. We specifically consider the case where the deformation model
allows for finite-element-type modeling of tissue properties. This
optimization problem is non-smooth, necessitating techniques like EAs to
get good results. Though the objectives of DIR are non-separable,
non-neighboring regions of the deformation grid are conditionally
independent. We show that GOMEA allows to exploit such knowledge through
the large-scale parallel application of variation steps, where each is
only accepted when leading to an improvement, on a Graphics Processing
Unit (GPU). On various 2-dimensional DIR problems, we find that this
way, similar results can be achieved as when sequential processing is
performed, while allowing for substantial speed-ups (up to a factor of
111) for the highest-dimensional problems (i.e., the highest
deformation-grid resolution). This work opens the door to the extension
of this type of DIR to larger (3-dimensional) deformation grids, and its
application to other real-world problems.},
keywords = {Deformable models, Image registration, Image resolution, Computational modeling, Graphics processing units, Evolutionary computation, Parallel processing, GOMEA, GPU parallelization, deformable image registration, multi-objective optimization},
doi = {doi:10.1109/CEC45853.2021.9504840},
  notes = {Also known as \cite{9504840}},
)


@INPROCEEDINGS(Bagherbeik:2021:CEC,
%xplor 25 Aug 2021
  author = {Mohammad Bagherbeik and Ali Sheikholeslami},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Caching and Vectorization Schemes to Accelerate Local Search
Algorithms for Assignment Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2549--2558},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Assignment Problems are a class of NP-hard combinatorial
optimization problems with a wide range of real-world applications such
as Vehicle Routing and FPGA Block Placement. Despite technological
advances, solvers that target Assignment Problems still require
significant computing resources and time, especially as problem sizes
grow. This paper introduces novel cost function formulations to leverage
vector processing elements in accelerating local search algorithms for
solving Quadratic Assignment and Semi-Assignment problems. We
incorporate these vectorization methods within a Parallel Tempering
framework to solve some of the most difficult known Quadratic Assignment
and Semi-Assignment Problems up to sizes of 729 integer variables and
show that this solver system can perform upwards of 300 times faster
than other state-of-the-art solvers. We then conduct experiments to
quantify the performance and scaling of these vectorization methods and
qualify their situational strengths and trade-offs for use in future
algorithms and hardware systems.},
keywords = {Vehicle routing, Evolutionary computation, Computer architecture, Benchmark testing, Search problems, Cost function, Hardware, Quadratic Assignment Problem, Local Search, Vector Processing, Parallel Tempering},
doi = {doi:10.1109/CEC45853.2021.9504700},
  notes = {Also known as \cite{9504700}},
)


@INPROCEEDINGS(Maqbool:2021:CEC,
%xplor 25 Aug 2021
  author = {Fahad Maqbool and Saad Razzaq and Asif Yar and Hajira Jabeen},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Large Scale Distributed Optimization using Apache Spark:
Distributed Scalable Shade-Bat (DistSSB)},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2559--2566},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Large Scale Global Optimization (LSGO) is interesting for its
applications in machine learning, e.g. in deep learning or knowledge
graph embeddings. Evolutionary algorithms (EA) have been found efficient
in solving complex optimization problems. However, the performance of
conventional EAs degrades with the increasing number of decision
variables due to the lack of scalability. This paper proposes a
scalable, parallel, distributed and hybrid EA named Distributed Scalable
Shade-BatScalable Shade Bat (DistSSB) to solve LSGO problems. DistSSB is
inspired by the exploration capability of the SHADE algorithm and
exploitation feature of the Bat algorithm (BA). To achieve scalability,
DistSSB is implemented using the popular distributed in-memory
framework, Apache Spark. DistSSB distributes its population into
multiple sub-populations using the island model. Each sub-population is
independently evolved using SHADE or BA. After the migration interval,
the best solutions are broadcasted employing the mesh topology. We have
compared the scalability, efficiency, and efficacy of DistSSB with
SHADE-ILS (CEC-2018 Winner) and GL-SHADE algorithm on the CEC-2013 LSGO
benchmark function suite.For most functions, DistSSB has obtained better
optimization results in lesser execution time as compared to SHADE-ILS
and GL-SHADE. We have tested and shown the scalability of DistSSB for up
to "one million" dimensions, whereas SHADE-ILS and GLSHADE fail to scale
up for larger problems.},
keywords = {Deep learning, Scalability, Sociology, Evolutionary computation, Cluster computing, Benchmark testing, Topology, LSGO, Apache Spark, Differential Evolution, SHADE-ILS, GL-SHADE, Evolutionary Computation, Distributed, Parallel, Scalable},
doi = {doi:10.1109/CEC45853.2021.9504853},
  notes = {Also known as \cite{9504853}},
)


@INPROCEEDINGS(Venkat:2021:CEC,
%xplor 25 Aug 2021
  author = {Rochan Avlur Venkat and Zakaria Oussalem and Arya Kumar Bhattacharya},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Training Convolutional Neural Networks with Differential
Evolution using Concurrent Task Apportioning on Hybrid CPU-GPU
Architectures},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2567--2576},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={The core algorithm for training of Artificial Neural Nets
(ANNs) continues to remain the back-propagation (BP) algorithm - to an
extent that it is now considered as a paradigm of Deep Learning (DL).
Many important facets of DL, like hierarchical construction of features
across layers in image recognition, vanishing gradients, etc., are taken
for granted without recognizing that these may implicitly be induced by
BP itself. Evolutionary Algorithms (EAs) perform global optimization in
contrast to localized gradient descent of BP. If used extensively for
ANN training, they can potentially disrupt these assumed facets of DL -
and construct alternative and interesting perspectives. But they are
severely constrained by the need for large computational resources, as
they work concurrently on a population of candidate solutions. The bulk
of processing occurs in the forward pass through the ANN of thousands of
data samples - which can be efficiently parallelized on GPUs. However,
the candidates themselves can be launched in small groups on different
CPU cores - by exploiting their natural concurrency. Here we explore the
possibility of launching training of ANNs with EAs on hybrid CPU-GPU
systems with candidates split across CPUs and samples across GPU
threads. We conduct a series of experiments from which we synthesize a
successful mechanism for orders-of-magnitude speedup of EAs through
efficient apportioning of multi-level computing tasks onto different
classes of processing elements. This enables analysis of DL using EAs
empowering alternative interpretations of the above facets.},
keywords = {Training, Deep learning, Concurrent computing, Image recognition, Sociology, Graphics processing units, Evolutionary computation, Artificial Neural Network, Differential Evolution, CPU, GPU, MPI, CUDA, Parallelism, Convolutional Neural Network},
doi = {doi:10.1109/CEC45853.2021.9504878},
  notes = {Also known as \cite{9504878}},
)


@INPROCEEDINGS(Meselhi:2021:CEC,
%xplor 25 Aug 2021
  author = {Mohamed A. Meselhi and Saber M. Elsayed and Ruhul A. Sarker and Daryl L. Essam},
booktitle={2021 IEEE Congress on Evolutionary Computation (CEC)},
title={Parallel Evolutionary Algorithm for {EEG} Optimization Problems},
year={2021},
  editor = {Yew-Soon Ong},
pages = {2577--2584},
  address = {Krakow, Poland},
  month = {28 June-1 July},
  ISBN13 = {978-1-7281-8393-0},
abstract={Big data optimization has become an important research topic
in many disciplines. These optimization problems involve a large volume
of data, from different sources, in different formats, that are
generated at a high speed. For example, in the healthcare sector,
electroencephalography (EEG), which is a method for monitoring brain
signals and typically used to diagnose neurological disorders, generates
a large amount of data which, however, is often captured with artifacts
added from non-brain sources. Evolutionary algorithms are considered one
of the most successful approaches for solving many such complex
optimization problems. In this paper, a differential evolution algorithm
is developed to remove artifacts from EEG signals of interest, by using
the parallel computing ability of a Graphics Processing Unit. Two levels
of parallelization, variable and individual, are implemented, with a
gradient-based local search and adaptive control parameters incorporated
in order to enhance a search's convergence. The proposed algorithm is
tested using six single objective problems from the 2015 big data
optimization competition problems with 1024, 3072 and 4864 decision
variables, as both noise-free and with white noise. The results
presented in this paper indicate that the proposed algorithm is capable
of achieving high-quality solutions, and is up to 374.7 faster than the
state-of-the-art algorithms.},
keywords = {Heuristic algorithms, Graphics processing units, Evolutionary computation, Big Data, White noise, Parallel processing, Electroencephalography, Big optimization problems, big data, differential evolution, EEG signals},
doi = {doi:10.1109/CEC45853.2021.9504925},
  notes = {Also known as \cite{9504925}},
)