Search Strategies for an Anytime Usage of the Branch and Prune Algorithm

Raphaël Chenouard (1) , Alexandre Goldsztejn (2) , Christophe Jermann (2)

1 IRCCyN - Institut de Recherche en Communications et en Cybernétique de Nantes
2 LINA - Laboratoire d'Informatique de Nantes Atlantique

When applied to numerical CSPs, the branch and prune algorithm (BPA) computes a sharp covering of the solution set. The BPA is therefore impractical when the solution set is large, typically when it has a dimension larger than four or five which is often met in underconstrained problems. The purpose of this paper is to present a new search tree exploration strategy for BPA that hybridizes depth-first and breadth-first searches. This search strategy allows the BPA discovering potential solutions in different areas of the search space in early stages of the exploration, hence allowing an anytime usage of the BPA. The merits of the proposed search strategy are experimentally evaluated.

Mots clés: Numerical CSP; anytime; branch and prune

Domaines: Intelligence artificielle [cs.AI]

Raphaël Chenouard, Alexandre Goldsztejn, Christophe Jermann. Search Strategies for an Anytime Usage of the Branch and Prune Algorithm. International Joint Conference on Artificial Intelligence, 2009, United States. p. 468-473. ⟨hal-00421506⟩