On the Generalization Ability of Geometric Semantic Genetic Programming

Created by W.Langdon from gp-bibliography.bib Revision:1.4524

  author =       "Ivo Goncalves and Sara Silva and Carlos M. Fonseca",
  title =        "On the Generalization Ability of Geometric Semantic
                 Genetic Programming",
  booktitle =    "18th European Conference on Genetic Programming",
  year =         "2015",
  editor =       "Penousal Machado and Malcolm I. Heywood and 
                 James McDermott and Mauro Castelli and 
                 Pablo Garcia-Sanchez and Paolo Burelli and Sebastian Risi and Kevin Sim",
  series =       "LNCS",
  volume =       "9025",
  publisher =    "Springer",
  pages =        "41--52",
  address =      "Copenhagen",
  month =        "8-10 " # apr,
  organisation = "EvoStar",
  keywords =     "genetic algorithms, genetic programming, Geometric
                 semantic genetic programming, Generalisation,
                 Overfitting, Pharmacokinetics, Drug discovery",
  isbn13 =       "978-3-319-16500-4",
  DOI =          "doi:10.1007/978-3-319-16501-1_4",
  abstract =     "Geometric Semantic Genetic Programming (GSGP) is a
                 recently proposed form of Genetic Programming (GP) that
                 searches directly the space of the underlying semantics
                 of the programs. The fitness landscape seen by the GSGP
                 variation operators is unimodal with a linear slope by
                 construction and, consequently, easy to search. Despite
                 this advantage, the offspring produced by these
                 operators grow very quickly. A new implementation of
                 the same operators was proposed that computes the
                 semantics of the offspring without having to explicitly
                 build their syntax. This allowed GSGP to be used for
                 the first time in real-life multidimensional datasets.
                 GSGP presented a surprisingly good generalisation
                 ability, which was justified by some properties of the
                 geometric semantic operators. In this paper, we show
                 that the good generalization ability of GSGP was the
                 result of a small implementation deviation from the
                 original formulation of the mutation operator, and that
                 without it the generalization results would be
                 significantly worse. We explain the reason for this
                 difference, and then we propose two variants of the
                 geometric semantic mutation that deterministically and
                 optimally adapt the mutation step. They reveal to be
                 more efficient in learning the training data, and they
                 also achieve a competitive generalization in only a
                 single operator application. This provides a
                 competitive alternative when performing semantic
                 search, particularly since they produce small
                 individuals and compute fast.",
  notes =        "Nominated for EuroGP 2015 Best Paper.

                 Part of \cite{Machado:2015:GP} EuroGP'2015 held in
                 conjunction with EvoCOP2015, EvoMusArt2015 and

Genetic Programming entries for Ivo Goncalves Sara Silva Carlos M Fonseca