Show simple item record

hal.structure.identifierLaboratoire d'analyse et modélisation de systèmes pour l'aide à la décision [LAMSADE]
dc.contributor.authorCazenave, Tristan
HAL ID: 743184
*
hal.structure.identifierLaboratoire d'Informatique Avancée de Saint-Denis [LIASD]
dc.contributor.authorMéhat, Jean*
hal.structure.identifierLaboratoire d'analyse et modélisation de systèmes pour l'aide à la décision [LAMSADE]
dc.contributor.authorSaffidine, Abdallah*
dc.date.accessioned2013-07-17T08:25:52Z
dc.date.available2013-07-17T08:25:52Z
dc.date.issued2012
dc.identifier.issn0950-7051
dc.identifier.urihttps://basepub.dauphine.fr/handle/123456789/11569
dc.language.isoenen
dc.subjectMonte-Carlo Tree Search
dc.subjectTranspositions
dc.subjectUCT Algorithm
dc.subjectDirect acyclic graph
dc.subjectGame tree search
dc.subjectHeuristic search
dc.subject.ddc006.3en
dc.titleUCD : Upper confidence bound for rooted directed acyclic graphs
dc.typeArticle accepté pour publication ou publié
dc.description.abstractenIn this paper we present a framework for testing various algorithms that deal with transpositions in Monte-Carlo Tree Search (MCTS). We call this framework Upper Confidence bound for Direct acyclic graphs (UCD) as it constitutes an extension of Upper Confidence bound for Trees (UCT) for Direct acyclic graphs (DAG).When using transpositions in MCTS, a DAG is progressively developed instead of a tree. There are multiple ways to handle the exploration exploitation dilemma when dealing with transpositions. We propose parameterized ways to compute the mean of the child, the playouts of the parent and the playouts of the child. We test the resulting algorithms on several games. For all games, original configurations of our algorithms improve on state of the art algorithms.
dc.relation.isversionofjnlnameKnowledge-Based Systems
dc.relation.isversionofjnlvol34
dc.relation.isversionofjnldate2012
dc.relation.isversionofjnlpages26-33
dc.relation.isversionofdoi10.1016/j.knosys.2011.11.014
dc.relation.isversionofjnlpublisherElsevier
dc.subject.ddclabelIntelligence artificielleen
dc.relation.forthcomingnonen
dc.relation.forthcomingprintnonen
dc.description.ssrncandidatenon
dc.description.halcandidateoui
dc.description.readershiprecherche
dc.description.audienceInternational
dc.relation.Isversionofjnlpeerreviewedoui
dc.date.updated2017-01-23T09:49:52Z
hal.identifierhal-01499672*
hal.version1*
hal.update.actionupdateMetadata*
hal.author.functionaut
hal.author.functionaut
hal.author.functionaut


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record