Accéder directement au contenu Accéder directement à la navigation
Communication dans un congrès

A dynamic island model for adaptive operator selection

Abstract :

In this paper we propose a generic framework for Dynamic Island Models, which can be used as an original approach for the adaptive selection of operators in evolutionary algorithms. Assigning a variation operator to each island, we show that the dynamic regulation of migrations, which takes into account the pertinence of recent migrations, distributes the individuals on the most promising islands, i.e., the most efficient operators, at each stage of the search. The efficiency of this approach is assessed on the One-Max problem by comparing theoretical expected results to those obtained by our dynamic island model. Experiments show that the model provides the expected behavior.

Type de document :
Communication dans un congrès
Liste complète des métadonnées

https://hal.univ-angers.fr/hal-03255428
Contributeur : Okina Université d'Angers <>
Soumis le : mercredi 9 juin 2021 - 15:13:52
Dernière modification le : jeudi 10 juin 2021 - 03:39:57

Identifiants

Collections

Citation

Caner Candan, Adrien Goëffon, Frédéric Lardeux, Frédéric Saubion. A dynamic island model for adaptive operator selection. 14th international conference on genetic and evolutionary computation conference, GECCO'12, 2012, Philadelphie, United States. pp.1253 - 1260, ⟨10.1145/2330163.2330337⟩. ⟨hal-03255428⟩

Partager

Métriques

Consultations de la notice

8