- author = "Miguel Nicolau",
- title = "Moving Towards Big Data Scalability with the Grammatical Evolution System",
- booktitle = "EURO 2015",
- year = "2015",
- address = "University of Strathclyde, Glasgow, UK",
- month = "12-15 " # jul,
- organisation = "the European Association of Operational Research Societies in conjunction with the UK OR Society",
- keywords = "genetic algorithms, genetic programming, Grammatical Evolution",
-
URL = "
https://researchrepository.ucd.ie/rest/bitstreams/24606/",
-
URL = "
https://files.core.ac.uk/download/pdf/43251904.pdf",
- size = "1 page?",
-
abstract = "Evolutionary Computation techniques in general, and
Genetic Programming (GP) type systems in particular,
are symbolic combination search algorithms, often used
in regression problems. The solutions these systems
provide are often competitive when compared to similar
learning systems, and, being in symbolic form, have the
potential to provide insight into relationships between
predictors, and estimate the overall complexity of
resulting models.
With the advent of big data and the increasing need of effective modeling, GP-like systems provide several advantages. These systems present a highly parallelisable structure, where each member of a population of semi-independent solution candidates is individually applied to a set of training samples.
The increasing presence of connected computing devices presents a formidable opportunity for the scalability of GP-like systems. In this work, we propose and partially implement a framework to deploy one such system, Grammatical Evolution, across a highly heterogeneous, asynchronous network of computing devices. We work towards a system combining the dynamic nature of such a network with the inherent adaptability of evolutionary systems. Early experiments are designed, using the open-source million song dataset.",
-
notes = "NB Euro 2015 euro2015@mci-group.com
https://euro2015.euro-online.org/
MA-70 Monday, 8:30-10:00 - Livingstone LT303, Level 3 Hyper-heuristics and Evolutionary Learning Stream: Data Science for Optimisation Invited session Chair: Daniel Karapetyan",
