Genetic Programming, Validation Sets, and Parsimony Pressure

Created by W.Langdon from gp-bibliography.bib Revision:1.8051

@TechReport{oai:hal.ccsd.cnrs.fr:inria-00000996_v1,

• title = "Genetic Programming, Validation Sets, and Parsimony Pressure",
• author = "Christian Gagn{\'e} and Marc Schoenauer and Marc Parizeau and Marco Tomassini",
• publisher = "HAL - CCSd - CNRS",
• year = "2006",
• month = jan # "~09",
• institution = "l'Equipe TAO INRIA Futurs",
• type = "ARTCOLLOQUE",
• number = "inria-00000996",
• address = "LRI Bat. 490, Universite Paris Sud, 91405 Orsay CEDEX, France",
• annote = "Christian Gagn{\'e} ",
• bibsource = "OAI-PMH server at hal.ccsd.cnrs.fr",
• contributor = "Christian Gagn{\'e} ",
• identifier = "inria-00000996 (version 1)",
• oai = "oai:hal.ccsd.cnrs.fr:inria-00000996_v1",
• keywords = "genetic algorithms, genetic programming, Computer Science/Learning",
• URL = "http://hal.inria.fr/inria-00000996/en/",
• URL = "http://hal.ccsd.cnrs.fr/docs/00/05/44/78/PDF/gagne-paper.pdf",
• URL = "http://arxiv.org/abs/cs/0601044",
• size = "12 pages",
• abstract = "Fitness functions based on test cases are very common in Genetic Programming (GP). This process can be assimilated to a learning task, with the inference of models from a limited number of samples. This paper is an investigation on two methods to improve generalization in GP-based learning: 1) the selection of the best-of-run individuals using a three data sets methodology, and 2) the application of parsimony pressure in order to reduce the complexity of the solutions. Results using GP in a binary classification setup show that while the accuracy on the test sets is preserved, with less variances compared to baseline results, the mean tree size obtained with the tested methods is significantly reduced.",
• notes = "See also \cite{eurogp06:GagneSchoenauerParizeauTomassini}",

}

Genetic Programming entries for Christian Gagne Marc Schoenauer Marc Parizeau Marco Tomassini

Citations