Differential Evolution for Strongly Noisy Optimization: Use 1.01$^n$ Resamplings at Iteration n and Reach the -1/2 Slope

Shih-Yuan Chiu 1 Ching-Nung Lin 1 Jialin Liu 2, 3 Tsang-Cheng Su 1 Fabien Teytaud 4 Olivier Teytaud 2, 3 Shi-Jim Yen 1
1 NDHU - National Dong Hwa University
2 TAO - Machine Learning and Optimisation
CNRS - Centre National de la Recherche Scientifique : UMR8623, Inria Saclay - Ile de France, UP11 - Université Paris-Sud - Paris 11, LRI - Laboratoire de Recherche en Informatique
3 LRI - Laboratoire de Recherche en Informatique
4 Université Lille Nord de France (COMUE)
Abstract : This paper is devoted to noisy optimization in case of a noise with standard deviation as large as variations of the fitness values, specifically when the variance does not decrease to zero around the optimum. We focus on comparing methods for choosing the number of resamplings. Experiments are performed on the differential evolution algorithm. By mathematical analysis, we design a new rule for choosing the number of resamplings for noisy optimization, as a function of the dimension, and validate its efficiency compared to existing heuristics.
Keywords : Noisy Optimization Resampling Differential Evolution
Type de document :
Communication dans un congrès
2015 IEEE Congress on Evolutionary Computation (IEEE CEC), May 2015, Sendai, Japan. 2015, 〈http://www.cec2015.org〉
Liste complète des métadonnées

Littérature citée [33 références]  Voir  Masquer  Télécharger
https://hal.inria.fr/hal-01120892
Contributeur : Jialin Liu <>
Soumis le : jeudi 26 février 2015 - 17:56:11
Dernière modification le : jeudi 21 février 2019 - 11:02:02
Document(s) archivé(s) le : mercredi 27 mai 2015 - 12:25:58

Fichier

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

Identifiants

  • HAL Id : hal-01120892, version 1

Collections

INSMI | UMR8623 | INRIA | LRI-AO | TDS-MACS | CENTRALESUPELEC | UNIV-PARIS-SACLAY

Citation

Shih-Yuan Chiu, Ching-Nung Lin, Jialin Liu, Tsang-Cheng Su, Fabien Teytaud, et al.. Differential Evolution for Strongly Noisy Optimization: Use 1.01$^n$ Resamplings at Iteration n and Reach the -1/2 Slope. 2015 IEEE Congress on Evolutionary Computation (IEEE CEC), May 2015, Sendai, Japan. 2015, 〈http://www.cec2015.org〉. 〈hal-01120892〉

Exporter

BibTeX TEI DC DCterms EndNote

Partager

Métriques

Consultations de la notice

468

Téléchargements de fichiers

581

Contact


archive-ouverte@inria.fr