%Created by Maarten Keijzer Mon, 05 Jul 2004 17:07:42 +0200 %GECCO http://www.isgec.org/ %WBL 15 Jan 2008 Fixup title etc. of chen:2004:lbp %WBL 20 Jul 2004 Add Keywords fixup authors etc @Proceedings{keijzer:2004:GECCO:lbp, title = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming, ACO, GEP, GNP", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/}, notes = {Distributed on CD-ROM at GECCO-2004}, } @InProceedings{banks:2004:lbp, author = "Edwin Roger Banks and James Hayes and Edwin Nunez", title = "Parametric Regression Through Genetic Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP001.pdf}, abstract = "Parametric regression in genetic programming can substantially speed up the search for solutions. Paradoxically, the same technique has difficulty finding a true optimum solution. The parametric formulation of a problem results in a fitness landscape that looks like an inverted brush with many bristles of almost equal length (individuals of high fitness), but with only one bristle that is very slightly longer than the rest, the optimum solution. As such it is easy to find very good, even outstanding solutions, but very difficult to locate the optimum solution. In this paper parametric regression is applied to a minimum-time-to-target problem. The solution is equivalent to the classical brachistochrone. Two formulations were tried: a parametric regression and the classical symbolic regression formulation. The parametric approach was superior without exception. We speculate the parametric approach is more generally applicable to other problems and suggest areas for more research.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{bagot:2004:lbp, author = "Benoit Bagot", title = "The Harmonic Decision Matrix: a group of operators for the fuzzy-logic, multi-objective decisions and optimizations ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP002.pdf}, abstract = "At the crossroad between fuzzy-logic, decision science, games theory, design of experiment, evolutionary algorithms and Pareto ranking, a new formula and its graphic representation helps to solve multi-objective problems; with possible applications in the development of complex technical systems, economics and many other situations. Used in a genetic program for the optimisation of an automatic gear box, this tool helps to conciliate numerous, partly opposing criteria, in order to emphasise a unique final solution.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{wilson:2004:lbp, author = "Garnet C. Wilson and Malcolm I. Heywood", title = "Search Operator Bias in Linearly Structured Genetic Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP003.pdf}, abstract = "GA solutions to the job-shop scheduling problem demonstrate that significant amounts of code context exist. Such observations have led to the introduction of biased search operators. In this work, we recognise that similar conditions exist in linearly structured GP (L-GP). An empirical study is made when biased search operators are applied to the San Mateo Trail (strategy), Two Box (regression), and Liver Disease (classification) benchmark problems. A preference is observed for biased mutation alone in the case of the regression problem, whereas the strategy and classification problems appear to prefer the combination of both biased mutation and crossover. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{jakob:2004:lbp, author = "Wilfried Jakob and Christian Blume and Georg Bretthauer", title = "Towards a Generally Applicable Self-adapting Hybridization of Evolutionary Algorithms ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP004.pdf}, abstract = "When applied to real-world problems, the powerful optimisation tool of Evolutionary Algorithms frequently turns out to be too time-consuming due to elaborate fitness calculations that are often based on run-time-intensive simulations. Incorporating domain-specific knowledge by problem-tailored heuristics or local searchers is a commonly used solution, but turns the generally applicable Evolutionary Algorithm into a problem-specific tool. The new method of hybridization implemented in HyGLEAM is aimed at overcoming this limitation and getting the best of both algorithm classes: A fast, globally searching, and robust procedure that preserves the convergence reliability of evolutionary search. Extensive tests demonstrate the superiority of the approach, but also show a drawback: No common parameterisation can be drawn from the experiments. As a solution, a new concept of a self-adapting hybrid is introduced. It is stressed that the methods presented can be applied to Evolutionary Algorithms other than the one used here with no or minor modifications being required only. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{uyar:2004:lbp, author = "A. Sima Uyar", title = "An Adaptive Diploid Evolutionary Algorithm for Floating-Point Representations in Dynamic Environments", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP005.pdf}, abstract = "Many flavours of different EA approaches have been suggested in literature to address various requirements that arise when the environment is dynamic. One such line of research focuses on using memory-based techniques. However, their use is very limited and they are only considered to be advantageous when the environment periodically returns to previously visited states. In addition, most of these approaches require a change detection mechanism to adapt and their performance partly depends on an accurate detection mechanism. In this study, an implicit memory-based approach for floating-point representations is introduced.. A technique similar to estimation of distribution algorithms is used to automatically adapt the population to the change. This approach aims at addressing some of the deficiencies reported for memory-based approaches and it requires no change detection mechanism to trigger the adaptation process. The Moving Peaks Benchmark is used in the experiments to show the desired properties of the proposed approach. Preliminary experimental results are promising and promote further study.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{wloch:2004:lbp, author = "Krzysztof Wloch and Peter J. Bentley", title = "Optimising the Performance of a Formula One Car using a Genetic Algorithm", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP006.pdf}, abstract = "Formula One motor racing is a rich sport that spends millions on research and development of highly optimised cars. Here we describe the use of a genetic algorithm to optimise 66 setup parameters for a simulation of a Formula One car and demonstrate performance improvements (faster lap times) better than all other methods tested.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{lee:2004:lbp, author = "Chungnan Lee and Yi-Te Li and Jain-Shing Wu and Ta-Yuan Chou", title = "Double Orthogonal Arrays Based Genetic Algorithm for Primer Design", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP007.pdf}, abstract = "In this paper, a double orthogonal array based genetic algorithm is proposed to decrease the searching space and increase the feasible quality of primers. The key point of the proposed algorithm is to achieve the elitism goal by applying the orthogonal arrays (OAs) that is used in quality engineering with a small amount of experimental features. The result shows that the proposed algorithm converges faster than the traditional GA. Based on the design properties such as unique, GC content, melting temperature, temperature difference and annealing, the proposed algorithm can derive the solutions more precisely and efficiently. The primer pairs obtained can exactly match to the PCR traits.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{wang:2004:lbp, author = "Z. G. Wang and Y. S. Wong and M. Rahman", title = "Development of the parallel optimization method based on genetic simulated annealing", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP008.pdf}, abstract = "This paper presents a parallel genetic simulated annealing (PGSA) algorithm that has been developed and applied to optimise continuous problems. In PGSA, the entire population is divided into subpopulations, and in each subpopulation the algorithm uses the local search ability of simulated annealing after crossover and mutation. The best individuals of each subpopulation are migrated to neighbouring ones after certain number of epochs. An implementation of the algorithm is discussed and the performance evaluation is made against a standard set of test functions. PGSA shows some remarkable improvement in comparison with the conventional simulated annealing, parallel genetic algorithm. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{piaseczny:2004:lbp, author = "Wojciech Piaseczny and Hideaki Suzuki and Hidefumi Sawai", title = "Chemical Genetic Programming - The Effect of Evolving Amino Acids", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP009.pdf}, abstract = "A new method of genetic programming, named chemical genetic programming (CGP), which enables evolutionary optimisation of the mapping from genotypic strings to phenotypic trees is proposed. A cell is evolved, and includes a DNA string that codes genetic information and smaller molecules for the mapping from DNA code to computational functionality. Genetic modification of a cell's DNA allows the DNA code and the genotype-to-phenotype translation to coevolve. Building an optimal translation table enhances evolution within a population while maintaining the necessary diversity to explore the entire search space.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{hsu:2004:lbp, author = "William H. Hsu and Scott J. Harmon and Edwin Rodriguez and Christopher Zhong", title = "Empirical Comparison of Incremental Reuse Strategies in Genetic Programming for Keep-Away Soccer", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP010.pdf}, abstract = "Easy missions approaches to machine learning seek to synthesise solutions for complex tasks from those for simpler ones. In genetic programming, this has been achieved by identifying goals and fitness functions for subproblems of the overall problem. Solutions evolved for these subproblems are then reused to speed up learning, either as automatically defined functions (ADFs) or by seeding a new GP population. Previous positive results using both approaches for learning in multi-agent systems (MAS) showed that incremental reuse using easy missions achieves comparable or better overall fitness than monolithic simple GP. A key unresolved issue dealt with hybrid reuse using ADF plus easy missions. Results in the keep-away soccer domain (a test bed for MAS learning) were also inconclusive on whether compactness inducing reuse helped or hurt overall agent performance. In this paper, we compare monolithic (simple GP and GP with ADFs) and easy missions reuse to two types of GP learning systems with incremental reuse: GP/ADF hybrids with easy missions and single-mission incremental ADFs. As hypothesised, pure easy missions reuse achieves results competitive with the best hybrid approaches in this domain. We interpret this finding and suggest a theoretical approach to characterising incremental reuse and code growth.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{barlow:2004:lbp, author = "Gregory J. Barlow and Choong K. Oh and Edward Grant", title = "Incremental Evolution of Autonomous Controllers for Unmanned Aerial Vehicles using Multi-objective Genetic Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP011.pdf}, abstract = "Autonomous navigation controllers were developed for fixed wing unmanned aerial vehicle (UAV) applications using incremental evolution with multi-objective genetic programming (GP). We designed four fitness functions derived from flight simulations and used multi-objective GP to evolve controllers able to locate a radar source, navigate the UAV to the source efficiently using on-board sensor measurements, and circle closely around the emitter. We selected realistic flight parameters and sensor inputs to aid in the transference of evolved controllers to physical UAVs. We used both direct and environmental incremental evolution to evolve controllers for four types of radars: 1) continuously emitting, stationary radars, 2) continuously emitting, mobile radars, 3) intermittently emitting, stationary radars, and 4) intermittently emitting, mobile radars. The use of incremental evolution drastically increased evolution's chances of evolving a successful controller compared to direct evolution. This technique can also be used to develop a single controller capable of handling all four radar types. In the next stage of research, the best evolved controllers will be tested by using them to fly real UAVs.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{zeng:2004:lbp, author = "Sanyou Zeng and Lixin Ding and Shuzhen Yao and Lishan Kang", title = "{KLP} Not Always Efficient", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP012.pdf}, abstract = "The size of the nondominated set of a vector set is greatly dependent on the size of the original vector set $N$ and the dimension of the vector $M$. Theoretical analysis shows that when $M=O(\log N)$ the original set has big nondominated set which may be the original set itself, and in the case $M=O(\log N)$, a classical algorithm (KLP) for finding nondominated set has complexity of KLP higher than $N^2$. Experiment verifies the analysis result as well. Therefore, we should avoid employing KLP when $M=O(\log N)$.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{lobo:2004:lbp, author = "Fernando G. Lobo", title = "A philosophical essay on life and its connections with genetic algorithms ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP013.pdf}, abstract = "This paper makes a number of connections between life and various facets of genetic and evolutionary algorithms research. Specifically, it addresses the topics of adaptation, multiobjective optimization, decision making, deception, and search operators, among others. It argues that human life, from birth to death, is an adaptive or dynamic optimisation problem where people are continuously searching for happiness. More important, the paper speculates that genetic algorithms can be used as a source of inspiration for helping people make decisions in their everyday life.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{lobo:2004:lbp-2, author = "Fernando G. Lobo and Claudio Lima and Hugo Martires", title = "An architecture for massive parallelization of the compact genetic algorithm ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP014.pdf}, abstract = "This paper presents an architecture which is suitable for a massive parallelisation of the compact genetic algorithm. The resulting scheme has three major advantages. First, it has low synchronisation costs. Second, it is fault tolerant, and third, it is scalable. The paper argues that the benefits that can be obtained with the proposed approach is potentially higher than those obtained with traditional parallel genetic algorithms. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{fernlund:2004:lbp, author = "Hans Fernlund and Avelino J. Gonzalez", title = "Using GP to Model Contextual Human Behavior - Competitive with Human Modeling Performance", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP015.pdf}, abstract = "To create a realistic environment, some simulations require simulated agents with human behaviour pattern. Creating such agents with realistic behavior can be a tedious and time consuming work. This paper describes a new approach that automatically builds human behaviour models for simulated agents by observing human performance. With an automatic tool that builds human behavioral agents, the development cost and effort could be dramatically reduced. This research synergistically combines Context-Based Reasoning (CxBR), a paradigm especially developed to model tactical human performance within simulated agents, with the Genetic Programming machine learning algorithm able to construct the behaviour knowledge in accordance to the CxBR paradigm. This synergistic combination of AI methodologies has resulted in a new algorithm that automatically builds simulated agents with human behavior. This algorithm was exhaustively tested with five different simulated agents created by observing the performance of five humans driving an automobile simulator. The agents show, not only the capabilities to automatically learn and generalise the behaviour of the human observed, but they also exhibited a performance that was at least as good as that of agents developed manually by a knowledge engineer. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{basanta:2004:lbp, author = "David Basanta and Mark Miodownik and Peter Bentley and Elizabeth Holm", title = "Investigating the evolvability of biologically inspired {CA}", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP016.pdf}, abstract = "The developmental processes studied by biologists are emergent self organised processes that are the result of natural evolution. Developmental biology can be a good inspiration for anyone interested in evolving self organised discrete systems like Cellular Automata: nature has proved that evolving self organisation is possible. In this paper, we will describe EmbryoCA, a model of 3D Cellular Automata for pattern generation. Developmental biology has been used as an inspiration to design EmbryoCA and make the model more gradual in the hope of having a more evolvable type of CA. Also, experiments comparing the evolvability of different setups of EmbryoCA with a conventional CA model are shown.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{poladian:2004:lbp, author = "Leon Poladian and Lars Jermiin", title = "Phylogenetic inference using evolutionary multi-objective optimisation", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP017.pdf}, abstract = "Evolutionary relationships among species are usually (i) illustrated by means of a phylogenetic tree and (ii) inferred by optimising some measure of fitness, such as the total evolutionary distance between species or the likelihood of the tree (given a model of the evolutionary process and a data set). The combinatorial complexity of inferring the topology of the best tree makes phylogenetic inference an ideal candidate for evolutionary algorithms. However, difficulties arise when different data sets provide conflicting information about the inferred `best' tree(s). We apply the techniques of multi-objective optimisation to phylogenetic inference. We present results for the simplest model of evolution and an artificially constructed four species problem. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{gang:2004:lbp, author = "Peng Gang and Ichiro Iimura and Hidenobu Tsurusawa and Shigeru Nakayama", title = "A Local Search Algorithm Based on Genetic Recombination for Traveling Salesman Problem", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP018.pdf}, abstract = "Genetic Algorithms(GAs) have been applied in many different fields and optimisation problem domains. It is well known that it is hard to solve the complex problems with a Simple Genetic Algorithm (SGA). Many previous studies have shown that the hybrid of local search and GAs is an effective approach for finding near optimum solutions to the travelling salesman problem (TSP). In this paper, an approach based on the Genetic Recombination is proposed and applied to the TSP. The algorithm is composed of two SGAs which only consist of the basic genetic operators such as selection, crossover and mutation. One of the SGAs is named as the Global Genetic Algorithm (GGA) and carried out in the main tours which are designed for searching the global optimal solutions. Another one is named as the Local Genetic Algorithm (LGA) and carried out in the sub tours which are designed for searching the local optimal solutions. The LGA is combined to the GGA as an operator. The local optimal solutions are recombined to the main tours for improving the search quality. To investigate the features of the proposed algorithm, it was applied to a small double circles TSP and some interesting results were presented in our experiments. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{james:2004:lbp, author = "Derek James and Philip Tucker", title = "A Comparative Analysis of Simplification and Complexification in the Evolution of Neural Network Topologies", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP019.pdf}, abstract = "Approaches to evolving the architectures of artificial neural networks have involved incrementally adding topological features (complexification), removing features (simplification), or both. We will present a comparative study of these dynamics, focusing on the domains of XOR and Tic-Tac-Toe, using NEAT (NeuroEvolution of Augmenting Topologies) as the starting point. Experimental comparisons are done using complexification, simplification, and a blend of both. Analysis of the effects of each approach on the variation, complexity, and fitness of the evolving populations demonstrates that algorithms employing both complexification and simplification dynamics search more efficiently and produce more compact solutions. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{Kasinadhuni:2004:lbp, author = "Maheswara Prasad Kasinadhuni and Michael L. Gargano and Joseph DeCicco and William Edelson", title = "SELF-ADAPTATION IN GENETIC ALGORITHMS USING MULTIPLE GENOMIC REDUNDANT REPRESENTATIONS", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP020.pdf}, abstract = "We consider an extension to optimisation problems [7, 13, 14] where the element costs are not fixed, but are time dependent. We propose using multiple genomic redundant representations in a self-adapting genetic algorithm (GA) employing various codes with different locality properties. These encoding schemes insure feasibility after performing the operations of crossover and mutation and also ensure the feasibility of the initial randomly generated population (i.e., generation 0). The GAs solving this class of NP hard problems, where costs are not fixed but are time dependent, employ non-locality or locality representations when appropriate (i.e., the GA adapts to its current search needs) which makes the GAs more efficient. A few applications with time dependent costs will also be presented. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{thangavelautham:2004:lbp, author = "Jekanthan Thangavelautham and Gabriele M. T. D'Eleuterio", title = "application of a Neuroevolutionary Approach to Emergent Task Decomposition in Collective Robotics", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP021.pdf}, abstract = "A scalable architecture to facilitate emergent (self-organised) task decomposition using neural networks and evolutionary algorithms is presented. Various control system architectures are compared for a collective robotics (tiling pattern formation) task where emergent behaviours and effective task-decomposition techniques are necessary to solve the task. We show that bigger, more modular network architectures that exploit emergent task decomposition strategies can evolve faster and outperform comparably smaller nonemergent neural networks for this task. Much like biological nervous systems, larger Emergent Task Decomposition Networks appear to evolve faster than comparable smaller networks. Unlike reinforcement learning techniques, only a global fitness function is specified, requiring limited supervision, and self-organised task decomposition is achieved through competition and specialisation. The results are derived from computer simulations. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{francone:2004:lbp, author = "Frank D. Francone and Larry M. Deschaine and Tom Battenhouse and Jeffrey J. Warren", title = "Discrimination of Unexploded Ordnance from Clutter Using Linear Genetic Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP022.pdf}, abstract = "We used Linear Genetic Programming (LGP) to study the extent to which automated learning techniques may be used to improve Unexploded Ordinance (UXO) discrimination from Protem-47 and Geonics EM61 non-invasive electromagnetic sensors. We conclude that: (1) Even after geophysicists have analysed the EM61 signals and ranked anomalies in order of the likelihood that each comprises UXO, our LGP tool was able to substantially improve the discrimination of UXO from scrap-preexisting techniques require digging 62% more holes to locate all UXO on a range than do LGP derived models; (2) LGP can improve discrimination even though trained on a very small number of examples of UXO; and (3) LGP can improve UXO discrimination on data sets that contain a high-level of noise and little preprocessing.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{li:2004:lbp, author = "Xin Li and Chi Zhou and Peter C. Nelson and Thomas M. Tirpak", title = "Investigation of Constant Creation Techniques in the Context of Gene Expression Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming, GEP", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP023.pdf}, abstract = "Gene Expression Programming (GEP) is a new technique of Genetic Programming (GP) that implements a linear genotype representation. It uses fixed-length chromosomes to represent expression trees of different shapes and sizes, which results in unconstrained search of the genome space while still ensuring validity of the programs output. However, GEP has some difficulty in discovering suitable function structures because the genetic operators are more disruptive than traditional tree-based GP. One possible remedy is to specifically assist the algorithm in discovering useful numeric constants. In this paper, the effectiveness of several constant creation techniques for GEP has been investigated through two symbolic regression benchmark problems. Our experimental results show that constant creation methods applied to the whole population for selected generations perform better than methods that are applied only to the best individuals. The proposed tune-up process for the entire population can significantly improve the average fitness of the best solutions.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{earon:2004:lbp, author = "E. J. P. Earon and G. M. T. D'Eleuterio", title = "An Agent Too Far: The Genetic Distance Evaluation of a Simulated World", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP024.pdf}, abstract = "One of the fundamental features, and one of the lesser understood phenomena, in biology is that of speciation. In order to better understand the development and creation of species, and their role in evolution, a method for tracking speciation in simulation is presented. While there is much dispute in the field of biology as to the precise definition of the term species, there is little debate that the natural world is full of distinct subpopulations. A genetic framework which permits a precise definition of a species is presented as it is implemented in the Cyberis simulator. The framework uses the notion of the Levenshtein distance as applied to the genomes of the agents to dynamically specify speciation during simulation. This allows for the study of artificial evolution on a more accurate scale by investigating the performance and longevity of the species, a more robust indicator of the genetics than a single agent. The genetics also allow much freedom in the specification of the agent and allows similar flexibility in experiments in neutrality and genetic difference calculations.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{skolicki:2004:lbp, author = "Zbigniew Skolicki and Kenneth {De Jong}", title = "Improving Evolutionary Algorithms with Multi-representation Island Models", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP025.pdf}, abstract = "We present an island model that uses different representations in each island. The model transforms individuals from one representation to another during migrations. We show that such a model helps the evolutionary algorithm to escape from local optima and to solve problems that are difficult for single representation EAs. We illustrate this approach with a two population island model in which one island uses a standard binary encoding and the other island uses a standard reflective Gray code. We compare the performance of this multi-representation island model with single population EAs using only binary or Gray codes. We show that, on a variety of difficult multi-modal test functions, the multi-representation island model does no worse than a standard EA on all of the functions, and produces significant improvements on a subset of them.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{lee:2004:lbp-2, author = "Kit-Ying Lee and Man-Leung Wong and Yong Liang and Kwong-Sak Leung and Kin-Hong Lee", title = "{A-HEP}: Adaptive Hybrid Evolutionary Programming for Learning Bayesian Networks", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP026.pdf}, abstract = "This paper describes an optimised algorithm for learning Bayesian Network structure by using adaptive population sized evolutionary programming. Bayesian network (BN) is a popular knowledge discovery model which can represent the causal relationship of different events or attributes with uncertainty. Learning the structure solely by dependency analysis or search-and-score approach is not effective. The hybrid algorithm on evolutionary programming, HEP, has been shown to be effective and efficient to solve this learning problem. By introducing the concept of adjusting the population size according to the individuals' dissimilarity, HEP is further optimised on the execution time with comparable performance. The empirical results illustrate that the optimized algorithm has reduced the running time by half.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{lefuel:2004:lbp, author = "Ramon Lefuel and Brian J. Ross", title = "Parsing Probabilistic Context Free Languages with Multi-Objective Genetic Algorithms", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP027.pdf}, abstract = "An approach to parsing probabilistic context free languages is presented. Given an input sentence, a genetic algorithm is used to evolve parse trees as defined by a given probabilistic context free grammar. Each chromosome in the population represents a candidate parse tree, using a simple indexed representation. The novelty of the approach is the multi-objective treatment of parse tree fitness. One dimension of the fitness space is the number of contiguous words correctly read by the parse. The other dimension incorporates a measurement equivalent to the probability obtained by complete parse trees, and partial probabilities corresponding to incomplete parses. A number of experiments show that this method is both effective and efficient for parsing natural language sentences.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{aldawoodi:2004:lbp, author = "Namir Aldawoodi and Rafael Perez and Wendy Alvis and Kimon Valavanis", title = "Developing Automated Helicopter Models Using Simulated Annealing and Genetic Search", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP029.pdf}, abstract = "A heuristic technique is presented that applies simulated annealing search to derive mathematical equations that model a pilot for an X-CELL 60 helicopter. The technique uses a pre-defined alphabet of formulas and combines them to create a mathematical model of the system controller or pilot. itself when the system is nonlinear and thus can not be modelled well with a PID. One such area where this method has great potential is in modeling the response of a human pilot flying a helicopter. A human response to the environment cannot be duplicated within a mathematical function that is normally used with PIDs this is due to the limited range of these functions.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{bagot:2004:lbp-2, author = "Benoit BAGOT and Hartmut POHLHEIM", title = "Complementary selection and variation for an efficient multiobjective optimization of complex systems ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP030.pdf}, abstract = "Real-world applications generally distinguish themselves from theoretical developments in that they are much more complex and varied. As a consequence, better models require more details, new methods and, finally, more complexity. By confronting a benchmark evolutionary algorithm with an automotive gearbox with hundreds of parameters to optimise, we were able to observe new requirements which led us to an additional procedure. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{lefort:2004:lbp, author = "Virginie LEFORT and Carole KNIBBE and Guillaume BESLON and Joel FAVREL", title = "The {RBF-Gene} model", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP031.pdf}, abstract = "We present here the Radial Basis Function Gene Model (RBF-Gene) as a new approach of evolutionary computation. This model introduces a quasi-realistic notion of gene into artificial genomes. Thus, it enables artificial organisms to evolve both on a functional basis (i.e. to enhance their fitness) and on a structural basis (i.e. to permanently reorganise their genome). This approach enables us to introduce new reproduction operators, thus giving the chromosome the opportunity to evolve its own `evolvability''. This allows a better convergence. In this paper, both principles and first results are presented, showing the great interest of this model. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{rotar:2004:lbp, author = "Corina Rotar", title = "An Evolutionary Technique for Multicriterial Optimization Based on Endocrine Paradigm", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP032.pdf}, abstract = "Many evolutionary algorithms have been lately developed for solving multiobjective problems, appealing or not to the Pareto optimality concept. Although, the evolutionary techniques for multiobjective optimization confront with several issues as: elitism, diversity of the population, or efficient settings for the specific parameters of the algorithm. In this paper, we propose a new evolutionary technique, which is inspired by the behavior of the endocrine system and uses the Pareto non-dominance concept. Therefore, the members of the population aren't called chromosomes anymore, but hormones and, even if they evolve according to the genetic principles, a supplementary mechanism based on the endocrine paradigm is connected with standard approach to deal with multiobjective optimization problems. Moreover, the proposed algorithm, in order to maintain the diversity of the population, uses a specific scheme of fit-ness sharing, eliminating the inconvenient of defining an appropriate value of sharing factor.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{stone:2004:lbp, author = "Sam Stone and Brian Pillmore and Walling Cyre", title = "Crossover and Mutation in Genetic Algorithms Using Graph-Encoded Chromosomes ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP033.pdf}, abstract = "Graph chromosomes provide an elegant and flexible structure whereby genetic algorithms can encode applications not easily represented by the conventional vector, list, or tree chromosomes. While general-purpose mutation operators for graph-encoded genetic algorithms are readily available, a graph-encoded GA also requires a general-purpose crossover operator that enables the GA to efficiently explore the search space. This paper describes the existing graph crossover operators and proposes a new crossover operator, GraphX. By operating on the graph's representation, rather than the graph's structure, the GraphX operator avoids the unnecessary complexities and performance penalties associated with the existing fragmentation/recombination operators. Experiments verify that the GraphX operator outperforms the traditional fragmentation/recombination operators, not only in terms of the fitness of the offspring, but also in terms of the amount of CPU time required to perform the crossover operation.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{grosan:2004:lbp, author = "Crina Grosan", title = "An Evolutionary Approach for Multiobjective Optimization using Adaptive Representation of Solutions ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP034.pdf}, abstract = "Many algorithms for multiobjective optimisation have been proposed in the last years. In the recent past a great importance have the MOEAs able to solve problems with more than two objectives and with a large number of decision vectors (space dimensions). The difficulties occur when problems with more than three objectives (higher dimensional problems) are considered. In this paper, a new algorithm for multiobjective optimization called Multiobjective Adaptive Representation Evolutionary Algorithm (MAREA) is proposed. MAREA combines an evolution strategy and an steady-state algorithm. The performance of the MAREA algorithm is assessed by using several well-known test functions having more than two objectives. MAREA is compared with the best present day algorithms: SPEA2, PESA and NSGA II. Results show that MAREA has a very good convergence. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{meyer:2004:lbp, author = "Bernd Meyer", title = "Convergence Control in {ACO}", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP035.pdf}, abstract = "Ant Colony Optimisation (ACO) is a recent stochastic meta-heuristic inspired by the foraging behaviour of real ants. As for all meta-heuristics the balance between learning based on previous solutions (intensification) and exploration of the search space (diversification) is of crucial importance. The present paper explores a novel approach to diversity control in ACO. The common idea of most diversity control mechanisms is to avoid or slow down full convergence. We suggest to instead use a fast converging search algorithm that is artificially confined to the critical phase of its convergence dynamics. We also analyse the influence of an ACO parameter that does not seem to have received sufficient attention in the ACO literature: alpha, the exponent on the pheromone level in the probabilistic choice function. Our studies suggest that $\alpha$ does not only qualitatively determine diversity and convergence behaviour, but also that a variable alpha can be used to render ant algorithms more robust. Based on these ideas we construct an Algorithm ccAS for which we present some encouraging results on standard benchmarks. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{mabu:2004:lbp, author = "Shingo Mabu and Kotaro Hirasawa and Jinglu Hu", title = "Genetic Network Programming with Reinforcement Learning and its Performance Evaluation ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming, GNP", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP036.pdf}, abstract = "A new graph-based evolutionary algorithm named 'Genetic Network Programming, GNP' has been proposed. GNP represents its solutions as graph structures, which can improve the expression ability and performance. Since GA, GP and GNP already proposed are based on evolution and they cannot change their solutions until one generation ends, we propose GNP with Reinforcement Learning (GNP with RL) in this paper in order to search solutions quickly. Evolutionary algorithm of GNP makes very compact graph structure which contributes to reducing the size of the Q-table and saving memory. Reinforcement Learning of GNP improves search speed for solutions because it can use the information obtained during task execution. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{monson:2004:lbp, author = "Christopher K. Monson and Kevin D. Seppi", title = "Improving on the Kalman Swarm: Extracting its Essential Characteristics", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP037.pdf}, abstract = "The Kalman Swarm (KSwarm) is a new approach to particle motion in PSO that reduces the number of iterations required to reach good solutions. Unfortunately, it has much higher computational complexity than basic PSO. This paper addresses the runtime of KSwarm in a new algorithm called ``Linear Kalman Swarm'' (LinkSwarm) which has linear complexity and performs even better than KSwarm. Some possible reasons for the success of KSwarm are also explored. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{cummins:2004:lbp, author = "Ronan Cummins and Colm O'Riordan", title = "Using Genetic Programming to Evolve Weighting Schemes for the Vector Space Model of Information Retrieval", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP038.pdf}, abstract = "Term weighting in many Information Retrieval models is of crucial importance in the research and development of accurate retrieval systems. This paper explores a method to automatically determine suitable term weighting schemes for the vector space model. Genetic Programming is used to automatically evolve weighting schemes that return a high average precision. These weighting functions are tested on well-known test collections and compared to the tf-idf based weighting scheme using standard Information Retrieval performance metrics.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{oberoi:2004:lbp, author = "Daman Oberoi and Bart Rylander", title = "Determining the Best Parent Selection Method for a Genetic Algorithm through Varying Problem Sizes and Complexities", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP039.pdf}, abstract = "We conduct an experiment to investigate which of three parent selection methods (elitism, fitness proportionate, and tournament) generates the greatest fitness in the fewest number of generations using a genetic algorithm (GA). The parent selection methods were applied to the problems of Maximum Ones, 3-Processor Scheduling, and Sorting, while in each case the problem size varied from 4 to 22. We show that for nearly all problem sizes and types, Tournament selection produces the best results. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{stone:2004:lbp-2, author = "Sam Stone and Brian Pillmore and Walling Cyre", title = "Crossover and Mutation in Genetic Algorithms Using Graph-Encoded Chromosomes", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP040.pdf}, abstract = "Graph chromosomes provide an elegant and flexible structure whereby genetic algorithms can encode applications not easily represented by the conventional vector, list, or tree chromosomes. While general-purpose mutation operators for graph-encoded genetic algorithms are readily available, a graph-encoded GA also requires a general-purpose crossover operator that enables the GA to efficiently explore the search space. This paper describes the existing graph crossover operators and proposes a new crossover operator, GraphX. By operating on the graph's representation, rather than the graph's structure, the GraphX operator avoids the unnecessary complexities and performance penalties associated with the existing fragmentation/recombination operators. Experiments verify that the GraphX operator outperforms the traditional fragmentation/recombination operators, not only in terms of the fitness of the offspring, but also in terms of the amount of CPU time required to perform the crossover operation. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{tchernev:2004:lbp, author = "Elko B. Tchernev and Dhananjay S. Phatak", title = "Control structures in linear and stack-based Genetic Programming ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP041.pdf}, abstract = "Genetic Programming, or GP, has traditionally used prefix trees for representation and reproduction, with implicit flow control. The different clauses (the evaluation condition, the if and else sections, etc.) are all subtrees of the flow-control node. Linear and stack-based representations, however, require explicit nodes to define the extent of the control structures. This paper introduces a stack-based technique for correct control structure creation and crossover, and discusses its implementation issues in linear and stack-based GP. A set of flow-control nodes is presented, and examples given for evolving an artificial ant on the Santa Fe and Los Altos Trails, with and without looping constructs.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{amin:2004:lbp, author = "Mohammad Amin and Malin Premaratne", title = "Constraint Handling of an Optical Components Selection Problem using a new Genetic Crossover Scheme", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP042.pdf}, abstract = "This paper proposes a new crossover scheme called Controlled Content Crossover (CCC). CCC is applied to solve a commercially important optimisation problem associated with the selection of the optimum set of components in optical network. CCC searches for the optimized solution and at the same time controls the value of the constrained parameter. This parameter is the optical component dispersion value in this paper. Therefore, by preserving the feasibility of the search points, CCC performs a genetic search over the feasible space. The simulation results show that the performance of the CCC is comparable with the answers obtained from CPLEX software and is better than the simple repair-based Genetic Algorithm proposed in the paper. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{doncieux:2004:lbp, author = "Stephane DONCIEUX and Samuel LANDAU and Nicolas GUELFI", title = "EcoSFERES: A Tool for the Design of Self-Organized Agent-Based Applications ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP043.pdf}, abstract = "In this article we present EcoSFERES, a framework that helps applying learning and evolutionary algorithms to the design of multi-agent systems (MAS). This framework includes tightly linked abstractions for evolutionary computing and multi-agent based simulations, and additionally, agent learning policies can also be included and reused straightforwardly. In particular, EcoSFERES makes it possible to implement a variety of learning and evolutionary techniques in order to compare their results in identical conditions, and it also facilitates important changes in the simulation setup. The main added values of this article are, on the one hand, to introduce the first development framework that provides a generic way of using evolutionary techniques for the design of MAS, and, on the other hand, to show that the tuning of MAS can be facilitated by a well designed development framework.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{murata:2004:lbp, author = "Tadahiko Murata and Takashi Nakamura", title = "Developing Cooperation of Multiple Agents Using Genetic Network Programming with Automatically Defined Groups ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming, GNP", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP044.pdf}, abstract = "In this paper, we propose Genetic Network Programming (GNP) Architecture using Automatically Defined Groups. GNP is a kind of new evolutionary method inspired from Genetic Programming (GP). While GP has a tree architecture, GNP has a network architecture, with which an agent works in the virtual world. Because only one network architecture is evolved for agents in a system in previous works, every agent takes actions in the same way. In this paper, we apply a coevolution model called Automatically Defined Groups (ADG) to an evolutionary process of GNP, so that several GNP architectures are evolved in order to develop a cooperation among multiple agents. By computer simulation, we show that multi-agent cooperation can be developed by our GNP architecture with the ADG model.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{kostikas:2004:lbp, author = "Konstantinos Kostikas and Charalambos Fragakis", title = "Genetic Programming for Guiding Branch and Bound Search", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP045.pdf}, abstract = "We propose how Genetic Programming (GP) can be used for developing, in real time, problem-specific heuristics for Branch and Bound (B&B) search. A GP run, embedded into the B&B process, exploits the characteristics of the particular problem being solved, evolving a problem-specific heuristic expression. The evolved heuristic replaces the default one for the rest of the B&B search. The application of our method to node selection for B&B based Mixed Integer Programming is illustrated by incorporating the GP node selection heuristic generator into a B&B MIP solver. The hybrid system compares well with the unmodified solver using DFS, BFS, or even the advanced Best Projection heuristic when confronted with hard MIP problems from the MIPLIB3 benchmarking suite.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{crawford-marks:2004:lbp, author = "Raphael Crawford-Marks and Lee Spector and Jon Klein", title = "Virtual Witches and Warlocks: A Quidditch Simulator and Quidditch-Playing Teams Coevolved via Genetic Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP046.pdf}, abstract = "Games make excellent challenge problems for Artificial Intelligence. Two-player turn-based games (Backgammon, Checkers, Chess) are easy to program, and AI players can be benchmarked against humans of varying skill levels. Recently, more complicated real-time team games have received attention because of their dynamic environments and the necessity for coordination. The RoboCup Soccer Simulator is the most popular and well-known of these environments. However, the soccer simulator is restricted to only two dimensions, and does not realistically model physics. In 2001, Spector et al. proposed creating a simulator of the imaginary game Quidditch from the Harry Potter Books by J.K. Rowling. This article describes such a simulator and the coevolved quidditch-playing teams created for it using Genetic Programming.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{rojas:2004:lbp, author = "Sergio A. Rojas and Peter J. Bentley", title = "A Grid-based Ant Colony System for Automatic Program Synthesis", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming, ACO", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP047.pdf}, abstract = "The Ant Colony Metaheuristic was originally proposed for tackling optimization problems. More recent research has suggested that it can be applied for automatic generation of programs. By allowing the artificial ants to visit functions and terminals nodes, they become able to build pheromone trails that represent computer programs for optimising a fitness domain-specific function. In this paper a novel approach is addressed using a grid architecture as a more suitable discrete world to be explored by the ants. The resulting system was applied to automatically produce programs to solve Boolean functions.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{langdon:2004:lbp, author = "W. B. Langdon and W. Banzhaf", title = "Repeated Sequences in Linear GP Genomes ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP048.pdf}, abstract = "Biological chromosomes are replete with repetitive sequences, microsatellites, SSR tracts, ALU, etc. in their DNA base sequences. We discover hierarchical repeating sequences (building blocks?) are evolved by genetic programming in linear time series prediction programs.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{pujol:2004:lbp, author = "Joao C. F. Pujol and Riccardo Poli", title = "A Highly Efficient Function Optimization with Genetic Programming ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP049.pdf}, abstract = "This paper describes a new approach for function optimisation that uses a novel representation for the parameters to be optimised. By using genetic programming using, the new method evolves functions that transform initial random values for the parameters into optimal ones. Moreover, the new approach addresses the scalability problem by using a representation that, in principle, is independent of the size of the problem being addressed. Promising results are reported, comparing the new method with differential evolution and particle swarm optimisation on a test suite of benchmark problems.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{cervone:2004:lbp, author = "Guido Cervone and Liviu Panait and Ramesh Singh and Menas Kafatos and Sean Luke", title = "An Application of Evolutionary Algorithms to Predict the Extent of {SLHF} Anomaly Associated with Coastal Earthquake ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP050.pdf}, abstract = "Multi sensor remote sensing provides real time high resolution data that can be used to study anomalous changes on land, in the ocean, and in the atmosphere associated with an impending earthquake. Anomalous behaviour in Surface Latent Heat Flux (SLHF) prior to large coastal earthquakes has been recently found. However, an SLHF time series usually contains several sharp peaks that may be associated either with earthquakes or with atmospheric perturbations. In this paper we have used evolutionary algorithms to perform a search in a large space bounded by longitude, latitude and time, to distinguish between signals associated with earthquakes and those associated with atmospheric phenomena. The algorithm finds paths which delimit the extent of the detected anomalies by optimising an objective function that takes into consideration several aspects, such as spatial and time continuity, the magnitude of the anomalies, and the distance to the continental boundary. This search strategy is crucial for the development of a fully automated early warning system for providing information about impending earthquakes in a seismically active coastal region. Experiments have been performed over a 2000 km^2 area comprising a part of the continental boundary between the African and Eurasian plate, roughly corresponding to Italy and Greece, one of the most seismically active regions. Using a 365-days-long time series, we identified three signals associated with seismic events. Additionally, it was possible to establish that the extent of the signal does not propagate further than 600 km from the epicenter of the earthquake.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{rodriguez-vazquez:2004:lbp, author = "Katya Rodriguez-Vazquez and Carlos Oliver-Morales", title = "Function Approximation by means of Multi-Branches Genetic Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP051.pdf}, abstract = "This work presents a performance analysis of a Multi-Branches Genetic Programming (MBGP) approach applied in symbolic regression (e.g. function approximation) problems. Genetic Programming (GP) has been previously applied to this kind of regression. However, one of the main drawbacks of GP is the fact that individuals tend to grow in size through the evolution process without a significant improvement in individual performance. In Multi-Branches Genetic Programming (MBGP), an individual is composed of several branches, each branch can evolve a part of individual solution, and final solution is composed of the integration of these partial solutions. Accurate solutions emerge by using MBGP consisting of a less complex structure in comparison with solutions generated by means of traditional GP encoding without considering any additional mechanisms such as a multi-objective fitness functions evaluation for tree size controlling.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{buehler:2004:lbp, author = "Erik Buehler and Sanjoy Das and Jack F. Cully", title = "Equilibrium and Extinction In a Trisexual Diploid Mating System", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP052.pdf}, abstract = "In order to study the dynamics of a three-sex (trisexual) mating system, we extend the heterogametic sex-determining mechanism, used in many species, to include three sexes: XX, XY and YY. In this model, non-like types may mate, but like-types may not mate. We compare the dynamics of this system to a Mendelian system under Hardy-Weinberg conditions, and coin the term Trisexual Equilibrium to describe a system state very similar to Hardy-Weinberg Equilibrium. We construct computer simulations and mathematical models in an attempt to quantify the system's dynamics, and conclude that three-sex systems are not stable over time; they are destined to converge to two-sex systems. This conclusion is based on the fact that the less-represented homozygote's frequency variance (between adjacent generations) is positively linearly proportional to the respective frequency its self. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{das:2004:lbp, author = "Sanjoy Das and Gurdip Singh and Sandeep Pujar and Praveen Koduru", title = "Ant Colony Algorithms for Routing in Sensor Networks", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP053.pdf}, abstract = "The ant systems optimisation approach is a new method of solving combinatorial optimization problems. It was originally introduced as a metaheuristic approach for the well-known travelling salesman problem. But it was subsequently shown to be an equally effective algorithm for solving other optimisation problems. In this paper, we present an ant colony algorithm for data-centric routing in sensor networks. In each pass of the proposed algorithm, ants are placed at the terminal nodes of the tree to be computed. They are then allowed to move towards one another, along the edges of the graph, until they merge into a single entity. In this process, the paths taken by the ants define a data distribution tree. Edges receive reinforcement in the form of pheromone deposits along the paths taken by the ant. Pheromones eventually accrue most along better edges. In addition to forward and backward ants, we also use random ants whose purpose is to enable sharing of information pertaining to a node potential with neighbouring sensors. Since ant algorithms perform computations solely through local interactions between ant-like agents by means of pheromones, they scale well for large-scale applications, and are particularly attractive for real world systems. The algorithm can easily be used in several practical applications. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{salazar:2004:lbp, author = "Daniel Salazar and Blas Galvan and Gabriel Winter", title = "Enhancing A Multiobjective Evolutionary Algorithm Through Flexible Evolution", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP054.pdf}, abstract = "In this paper the use of a powerful single-objective optimisation methodology in Multi-objective Optimization Algorithms (MOEAs) is introduced. The Flexible Evolution concepts (FE) have been recently developed and proved its efficiency gains compared with several Evolutionary Algorithms solving single-objective challenging problems. The main feature of such concepts is the flexibility to self-adapt the internal behaviour of the algorithm to optimise its search capacity. In this paper we present the first attempt to incorporate FE into MOEAs. A real coded NSGA-II algorithm was modified replacing the crossover and mutation operators with the Sampling Engine of FE. Other two FE characteristics were implemented too: The Probabilistic Control Mechanism and the Enlarged Individual s Code. The performance of the resulting algorithm has been compared with the classical NSGA-II using several test functions. The results obtained and presented show that FE_based algorithms have advantages over the classical ones, especially when optimising highly multimodal complex functions.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{woodward:2004:lbp, author = "John Woodward", title = "Simple Incremental Testing", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP055.pdf}, abstract = "Additional test cases are added one by one, as the population solves the current set of test cases, until some fixed final limit is reached. We examine the general nature of this approach. Complexity is defined in a general sense. It is proved that adding a test case to a test set never reduces the complexity of a solution, and never increases the probability of finding a solution. The terms representative and redundant, are formally defined. The variation in the number of test cases and the jumps in the number of test cases are observed. The size of the test set just before a general solution is found, indicates a threshold number of test cases required for generalisation. We observe, how generalization varies with the size of the test set. Finally we observe the number of successes per evaluation required to produce a general solution.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{khosraviani:2004:lbp, author = "Bijan KHosraviani and Raymond E. Levitt and John R. Koza", title = "Organization Design Optimization Using Genetic Programming ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP056.pdf}, abstract = "This paper describes how we use Genetic Programming (GP) techniques to help project managers find near optimal designs for their project organisations. We use GP as a postprocessor optimiser for the project organisation design simulator Virtual Design Team (VDT). Decision making policy and individual/sub-team properties, activity assignments and percentage allocation for each activity are varied by GP, and the effect on quality and duration of the project is compared via a fitness function. The solutions found by GP compare favourably with the bes", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{simske:2004:lbp, author = "Steven J. Simske and David C. Matthews", title = "Navigation Using Inverting Genetic Algorithms: Initial Conditions and Node-Node Transitions", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP057.pdf}, abstract = "Navigation path optimisation derives its cost function from the total cost of travel. These costs can accrue from distance, traffic patterns, preferred order of node sequencing, maximum preferred distance between nodes, and other pragmatic considerations (node-node costs). For a travelling salesman problem (TSP), these costs are usually distances. This paper considers the effects of the initial states in the 'genome' - focusing on the use of rule-based and clustering techniques for initial conditions. It also considers the effects of weighting the node-node transition costs on algorithm convergence and final path cost. The best tested combination of initial states and rules for recombination involves weighting each by the distances between nodes. In addition to the mean of the residual error m, the metric designated algorithmic efficacy (AE) is introduced as a useful comparative metric for navigational optimisation algorithms. Four novel and six published TSP problems are investigated: m is shown to improve in every case, while AE has a wide range of results. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{settles:2004:lbp, author = "Matthew Settles and Terence Soule", title = "Breeding Swarms: A {GA/PSO} Hybrid", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP058.pdf}, abstract = "In this paper we propose a novel hybrid (GA/PSO) algorithm, Breeding Swarm, combining the strengths of particle swarm optimisation with genetic algorithms. The hybrid algorithm combines the standard velocity and update rules of PSOs with the ideas of selection, crossover and mutation from GAs. We propose a new crossover operator (VPAC), incorporating the PSO velocity vector, which actively disperses the population preventing premature convergence. We compare the hybrid algorithm to both the standard GA and PSO models in evolving solutions to four standard function minimisation problems. Results show the algorithm to be highly competitive, often outperforming both the GA and PSO. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{khalifa:2004:lbp, author = "Yaser M. A. Khalifa and Hunter Shi and Gustavo Abreu", title = "Evolutionary Music Composer ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP059.pdf}, abstract = "In this paper, an autonomous intelligent music composition tool was developed using Genetic Algorithms. The research has been structured into two phases, each of which builds upon the previous one. The first phase of the project was to develop more sophisticated fitness measures for the genetic algorithm, with the goal of applying data compression techniques to identify musically sound patterns in music through music theory principles. In the second phase, methods to use weighted permutations of different fitness functions and generated motifs were explored. These combinations were evaluated and as a result, musically fit patterns were generated. Four musical phrases are generated at the end of each program run, each phrase consists of eight measures, and each measure is one motif of up to eight notes. The generated music piece will be translated through an additional algorithm into Guido Music Notation (GMN) files for further evaluation and alternate representation (midi). The Evolutionary Music Composer (EMC) was able to create interesting pieces of music that were both innovative and musically sound.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{vandecasteele:2004:lbp, author = "Frederik P. J. Vandecasteele and Thomas F. Hess and Ronald L. Crawford", title = "A Correlated Fitness Landscape Describes Growth in Experimental Microbial Ecosystems: Initial Results", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP060.pdf}, abstract = "This paper describes experimental tests of the biological notion that correlated fitness landscapes underlie the functioning of ecosystems. We have demonstrated that a correlated fitness landscape topology describes overall growth in microbial ecosystems as a function of species composition. This confirms the intuitive notion that the more similar the structures of ecosystems are, the more similar their functions will be. These methods and results provide opportunities for better understanding how ecosystems function. They also provide a rationale for using evolutionary computation in optimising microbial ecosystems.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{chia:2004:lbp, author = "Henry Wai-Kit Chia and Chew-Lim Tan", title = "Association-Based Evolution of Comprehensible Neural Logic Networks ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP061.pdf}, abstract = "Neural Logic Network (Neulonet) learning has been successfully used in emulating complex human reasoning processes. One recent implementation generates a single large neulonet via genetic programming using an accuracy-based fitness measure. However, in terms of human comprehensibility and amenability during logic inference, evolving multiple compact neulonets are preferred. The present work realizes this by adopting associative-classification measures of confidence and support as part of the fitness computation. The evolved neulonets are combined together to form an eventual macro-classier. Empirical study shows that associative classification integrated with neulonet learning performs better than general association-based classifiers in terms of higher accuracies and smaller rule sets. This is primarily due to the richness in logic expression inherent in the neulonet learning paradigm. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{chen:2004:lbp, author = "Shu-Heng Chen and Bin-Tzong Chie", title = "Functional Modularity in the Test Bed of Economic Theory -- Using Genetic Programming", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP062.pdf}, abstract = "In this paper, we follow the model of Chen and Chie (2004), but start with the primeval setup. The implementation of computer simulations show mutation did play an important role in the technology evolution. In a well define simulation world, the producer will exert all of his effort to make the life get better. The parameter of mutation rate is just like the frequency of innovation in the real world. Different mutation rate will shift the model to the different path of history. The path of real world might be represented by one of the mutation rate, but it must be emergent from the different behaviours of the bottom actors.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{lipson:2004:lbp, author = "Hod Lipson", title = "How to Draw a Straight Line Using a {GP}: Benchmarking Evolutionary Design Against 19th Century Kinematic Synthesis", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms, genetic programming", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP063.pdf}, abstract = "This paper discusses the application of genetic programming to the synthesis of compound 2D kinematic mechanisms, and benchmarks the results against one of the classical kinematic challenges of 19th century mechanical design. Considerations for selecting a representation for mechanism design are presented, and a number of human-competitive inventions are shown. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{khalifa:2004:lbp-2, author = "Yaser Khalifa and Ehi Okoene", title = "An Autonomous Agent-Based Surveillance System", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP064.pdf}, abstract = "In this paper, we present an evolutionary technique for an agent-based surveillance and target tracking system. The proposed system is composed of two stages. The first stage is the optimisation of a network of ultrasonic sensors within a pre-specified coverage area. In phase two, the target tracking phase, sensors are activated based on need according to the tracking requirement of the moving object. Unneeded sensors are kept in a low-power stand-by state to optimise power usage and hence minimise human intervention for maintenance purpose. A centralised server communicates with sensor agents to power-up those needed for proper tracking of the moving object. The system is designed in C++ and simulated using MATLAB.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{zykov:2004:lbp, author = "Viktor Zykov and Josh Bongard and Hod Lipson", title = "Evolving Dynamic Gaits on a Physical Robot", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP065.pdf}, abstract = "Here we introduce a method for the evolution of dynamic gaits on a physical robot requiring no prior assumptions about the locomotion pattern beyond the fact that it should be rhythmic. The dynamic gaits were physically evolved in hardware using a parallel-actuated pneumatic robot. We have formulated a genetic algorithm that evolves open-loop controllers; the encoding allows evolution to shape both the speed and pattern of locomotion while ensuring rhythmicity. In future, we plan to evolve closed-loop controllers for the physical robot and integrate our previously developed methods to reduce the number of hardware trials.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{gomez:2004:lbp, author = "Osvaldo Gomez and Benjamin Baran", title = "Relationship between Genetic Algorithms and Ant Colony Optimization", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP066.pdf}, abstract = "Genetic Algorithms (GAs) were introduced by Holland as a computational analogy of adaptive systems. GAs are search procedures based on the mechanics of natural selection and natural genetics. Ant Colony Optimisation (ACO) is a metaheuristic inspired by the foraging behaviour of ant colonies. ACO was introduced by Dorigo and has evolved significantly in the last few years. Both algorithms have shown their effectiveness in the resolution of hard combinatorial optimisation problems. This paper shows the relationship between these two evolutionary algorithms inspired by different nature phenomena.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{popovici:2004:lbp, author = "Elena Popovici and Kenneth {De Jong}", title = "Understanding Competitive Co-evolutionary Dynamics via Fitness Landscapes ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP067.pdf}, abstract = "Co-evolutionary EAs are often applied to optimisation and machine learning problems with disappointing results. One of the contributing factors to this is the complexity of the dynamics exhibited by co-evolutionary systems. In this paper we focus on a particular form of competitive co-evolutionary EA and study the dynamics of the fitness of the best individuals in the evolving populations. Our approach is to try to understand the characteristics of the fitness landscapes that produce particular kinds of fitness dynamics such as stable fixed points, stable cycles, and instability. In particular, we show how landscapes can be constructed that produce each of these dynamics. These landscapes are extremely similar when inspected with respect to traditional properties such as ruggedness/modality, yet they yield very different results. This shows there is a need for co-evolutionary specific analysis tools.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{kaige:2004:lbp, author = "Shiori Kaige and Kaname Narukawa and Hisao Ishibuchi", title = "Lamarckian Repair and Darwinian Repair in {EMO} Algorithms for Multiobjective 0/1 Knapsack Problems", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP068.pdf}, abstract = "Multiobjective 0/1 knapsack problems have been frequently used as test problems to examine the performance of evolutionary multiobjective optimisation algorithms in the literature. It has been reported that their performance strongly depends on the choice of a constraint handling method. In this paper, we examine two implementation schemes of greedy repair: Lamarckian and Darwinian. In the Lamarckian implementation of greedy repair, a feasible solution is generated from an infeasible one by removing items until all the constraint conditions are satisfied. That is, the genetic information of the unfeasible solution is modified. On the other hand, the genetic information of the infeasible solution is not changed in the Darwinian implementation where greedy repair is used only to evaluate the fitness value of each solution. We compare these two implementation schemes with each other through computational experiments. We also compare greedy repair-based methods with a penalty function approach.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{kazadi:2004:lbp, author = "Sanza Kazadi and Daniel Johnson and Jhanisus Melendez and Brian Goo", title = "Exhaustive Directed Search", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP069.pdf}, abstract = "We explore the development of an exhaustive directed search of state space based on concepts from evolutionary computation. A brief investigation of the evolvability of an evolutionary algorithm illustrates that evolutionary algorithms are capable of reaching optimal solutions when the diversification operator (which may be a \emph{pseudo-operator} which acts over many different diversification steps) is capable of reaching, at every improvement point, another, more improved population element. Moreover, we demonstrate that the upper limit on the time to the optimal point is identical to that of an \emph{exhaustive directed search}. This search is exhaustive, but borrows the diversification operator from the evolutionary algorithm and proceeds in such a way that, if left alone, it would exhaustively search the space. However, we demonstrate that this type of search can perform comparably with the evolutionary algorithm, avoiding deceptive search tracks that might trap an evolutionary algorithm.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{ando:2004:lbp, author = "Shin Ando and Shigenobu Kobayashi", title = "On the sampling property of real-parameter crossover", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP070.pdf}, abstract = "In solving real-world optimisation problems using evolutionary algorithms (EAs), researchers have recently developed a number of real-parameter genetic algorithms (GAs). In these studies, the main focus of the research is the recombination operator, which use probability distributions calculated from the parent solutions. Such operator is intrinsically a non-parametric modeller and estimator, which the success of the real-parameter GA depends on. In this paper, we propose an implementation of a graph structure among the population to select such non-parametric kernels more efficiently during the search process. By recording the successful crossover as an edge between the individuals, clusters within the evolved population were observed. Additionally, a simple experiment exploiting such clusters showed efficient result in multimodal optimisation.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{aldwoodi:2004:lbp, author = "Namir Aldwoodi and Rafael Perez", title = "Advanced Formula Prediction using Simulated Annealing", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP071.pdf}, abstract = "An improved heuristic technique based on a method that we developed in 2003 [1]. The original method was called Formula Prediction using Genetic Algorithms (FPEG) and that method presented an algorithm by which formulas could be generated directly from datasets. The method presented enhanced the power and the flexibility of the original model resulting a better formula generation tool. The enhanced method differs from the original by the formula structure and expressive power - it has a larger alphabet and new independent internal variables that allow for better data pattern recognition and better overall performance. The method presented here uses simulated annealing to generate mathematical equations that fit a set of input data to a function. Data is represented as a set of input and output values collected from a system under consideration. For significantly large numbers of independent variables in the input set, this problem can be intractable and as such NP hard. When Advanced Formula Prediction using Simulated Annealing (AFP) was compared against FPEG using the original benchmarks - the results obtained show that the new algorithm is able to better the original algorithm's performance by 2.65 percentage points on average - or - an average reduction of the error margin by 52.10 percent (which is statistically significant). To keep the comparison valid, the same regression benchmarks were used. In addition, a technique that encodes strings that represent the candidate formulas during the search was enhanced to give it more expressive power. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{kumar:2004:lbp, author = "Sanjeev Kumar", title = "The Evolution of Genetic Regulatory Networks for Single and Multicellular Development ", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP072.pdf}, abstract = "Recently, researchers have recognised the benefits of learning from biological development in order to engineer self-organising solutions to problems. Building upon previous work, this paper explores the application of the developmental metaphor to the problem of controlling single and multicellular development. In this paper, a summary of experiments performed using a multicellular test-bed model of biological development, the Evolutionary Developmental System (EDS), is presented. The EDS is shown to successfully evolve genetic regulatory networks that specify and control the behaviour of single cells and the construction of 3D multicellular geometric morphologies to explore self-organisation and phenomena akin to biological cell differentiation in multicellular development. ", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{citi:2004:lbp, author = "Luca Citi and Riccardo Poli and Caterina Cinel and Francisco Sepulveda", title = "Feature Selection and Classification in Brain Computer Interfaces by a Genetic Algorithm", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP073.pdf}, abstract = "In this paper we explore the use of evolutionary algorithms in a wrapper-based selection of features and the classification of P300 signals in Brain Computer Interfaces. In particular we focus on a paradigm that uses the P300 potential associated to particular visual stimuli for hands free text entering. In our experiments the GA has found new ways to process and combine EEG signals to improve P300 detection accuracy.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{timm:2004:lbp, author = "Richard W. Timm and Hod Lipson", title = "Periodicity Emerges from Evolved Energy-Efficient and Long-Range Brachiation", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP074.pdf}, abstract = "This paper examines the emergence of periodic motion as consequence of evolutionary pressure for energy-efficient partially-passive dynamic brachiation. Brachiation was simulated for gibbon-like creatures in a realistic physics environment. The morphologies of the brachiators were identical to each other and consisted of a simple 3-bodied system - two arms and a torso. Brachiator performance was controlled by their genome, which contained their initial position at t=0 and torque functions which were applied to the shoulder for t>0. Performance for each brachiator was measured by total distance travelled and energy-efficiency of locomotion. Brachiators found in the first two Pareto-optimum curves were selected for reproduction. It was found that energy-efficient long-range motion resulted in periodic motion in the brachiators.", notes = {Part of keijzer:2004:GECCO:lbp}, } @InProceedings{holmes:2004:lbp, author = "John H. Holmes and Jennifer A. Sager and Warren B. Bilker", title = "Methods for Covering Missing Data in {XCS}", booktitle = "Late Breaking Papers at the 2004 Genetic and Evolutionary Computation Conference", year = "2004", editor = "Maarten Keijzer", address = "Seattle, Washington, USA", month = "26 " # jul, keywords = "genetic algorithms", url = {http://www.cs.bham.ac.uk/~wbl/biblio/gecco2004/LBP075.pdf}, abstract = "Missing data pose a potential threat to learning and classification in that they may compromise the ability of a system to develop robust, generalised models of the environment in which they operate. This investigation reports on the effects of three approaches to covering these data using an XCS- style learning classifier system. Using fabricated datasets representing a wide range of missing value densities, it was found that missing data do not appear to adversely affect LCS learning and classification performance. Furthermore, three types of missing value covering were found to exhibit similar efficiency on these data, with respect to learning rate and classification accuracy.", notes = {Part of keijzer:2004:GECCO:lbp}, }