A comparison of predictive measures of problem difficulty for classification with Genetic Programming

Leonardo Trujillo 1 Yuliana Martinez 1 Edgar Galvan-Lopez 2 Pierrick Legrand 3, 4
2 School of Computer Science and Electronic Engineering
CSEE - School of Computer Science and Electronic Engineering [Essex]
4 ALEA - Advanced Learning Evolutionary Algorithms
Inria Bordeaux - Sud-Ouest, UB - Université de Bordeaux, CNRS - Centre National de la Recherche Scientifique : UMR5251
Abstract : In the field of Genetic Programming (GP) a question exists that is difficult to solve; how can problem difficulty be determined? In this paper the overall goal is to develop predictive tools that estimate how difficult a problem is for GP to solve. Here we analyse two groups of methods. We call the first group Evolvability Indicators (EI), measures that capture how amendable the fitness landscape is to a GP search. The second are Predictors of Expected Performance (PEP), models that take as input a set of descriptive attributes of a problem and predict the expected performance of a GP system. These predictive variables are domain specific thus problems are described in the context of the problem domain. This paper compares an EI, the Negative Slope Coefficient, and a PEP model for a GP classifier. Results suggest that the EI does not correlate with the performance of GP classifiers. Conversely, the PEP models show a high correlation with GP performance. It appears that while an EI estimates the difficulty of a search, it does not necessarily capture the difficulty of the underlying problem. However, while PEP models treat GP as a computational black-box, they can produce accurate performance predictions.
Type de document :
Communication dans un congrès
ERA 2012, Nov 2012, Tijuana, Mexico. 2012
Liste complète des métadonnées

Littérature citée [25 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-00757363
Contributeur : Pierrick Legrand <>
Soumis le : lundi 26 novembre 2012 - 16:54:47
Dernière modification le : jeudi 11 janvier 2018 - 06:22:36
Document(s) archivé(s) le : mercredi 27 février 2013 - 03:46:47

Fichier

ERA_2012_NSC.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-00757363, version 1

Collections

Citation

Leonardo Trujillo, Yuliana Martinez, Edgar Galvan-Lopez, Pierrick Legrand. A comparison of predictive measures of problem difficulty for classification with Genetic Programming. ERA 2012, Nov 2012, Tijuana, Mexico. 2012. 〈hal-00757363〉

Partager

Métriques

Consultations de la notice

497

Téléchargements de fichiers

188