Reinforcement learning based local search for grouping problems: A case study on graph coloring - Université d'Angers Accéder directement au contenu
Article Dans Une Revue Expert Systems with Applications Année : 2016

Reinforcement learning based local search for grouping problems: A case study on graph coloring

Résumé

Grouping problems aim to partition a set of items into multiple mutually disjoint subsets according to some specific criterion and constraints. Grouping problems cover a large class of computational problems including clustering and classification that frequently appear in expert and intelligent systems as well as many real applications. This paper focuses on developing a general-purpose solution approach for grouping problems, i.e., reinforcement learning based local search (RLS), which combines reinforcement learning techniques with local search. This paper makes the following contributions: we show that (1) reinforcement learning can help obtain useful information from discovered local optimum solutions; (2) the learned information can be advantageously used to guide the search algorithm towards promising regions. To the best of our knowledge, this is the first attempt to propose a formal model that combines reinforcement learning and local search for solving grouping problems. The proposed approach is verified on a well-known representative grouping problem (graph coloring). The generality of the approach makes it applicable to other grouping problems.
Fichier principal
Vignette du fichier
zhouduvalhaoeswa2016.pdf (299.07 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01412526 , version 1 (08-12-2016)
hal-01412526 , version 2 (04-06-2017)

Identifiants

Citer

Yangming Zhou, Jin-Kao Hao, Béatrice Duval. Reinforcement learning based local search for grouping problems: A case study on graph coloring. Expert Systems with Applications, 2016, 64, pp.412-422. ⟨10.1016/j.eswa.2016.07.047⟩. ⟨hal-01412526v2⟩

Collections

UNIV-ANGERS LERIA
45 Consultations
299 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More