Show simple item record

Best Arm Identification in Graphical Bilinear Bandits

dc.contributor.authorRizk, Geovani
dc.contributor.authorThomas , A.
dc.contributor.authorColin, Igor
dc.contributor.authorLaraki, Rida
HAL ID: 179670
ORCID: 0000-0002-4898-2424
dc.contributor.authorChevaleyre, Yann
dc.date.accessioned2022-02-28T08:57:45Z
dc.date.available2022-02-28T08:57:45Z
dc.date.issued2021
dc.identifier.urihttps://basepub.dauphine.psl.eu/handle/123456789/22780
dc.language.isoenen
dc.subjectgraphical bilinear bandit
dc.subject.ddc004en
dc.titleBest Arm Identification in Graphical Bilinear Bandits
dc.typeCommunication / Conférence
dc.description.abstractenWe introduce a new graphical bilinear bandit problem where a learner (or a \emph{central entity}) allocates arms to the nodes of a graph and observes for each edge a noisy bilinear reward representing the interaction between the two end nodes. We study the best arm identification problem in which the learner wants to find the graph allocation maximizing the sum of the bilinear rewards. By efficiently exploiting the geometry of this bandit problem, we propose a \emph{decentralized} allocation strategy based on random sampling with theoretical guarantees. In particular, we characterize the influence of the graph structure (e.g. star, complete or circle) on the convergence rate and propose empirical experiments that confirm this dependency.
dc.identifier.citationpages139:9010-9019
dc.relation.ispartofpublnameProceedings of the 38th International Conference on Machine Learning
dc.subject.ddclabelInformatique généraleen
dc.relation.confdate2021
dc.relation.confcountryUNITED STATES
dc.relation.forthcomingnonen
dc.description.ssrncandidatenon
dc.description.halcandidatenon
dc.description.readershiprecherche
dc.description.audienceInternational
dc.date.updated2022-06-19T22:42:35Z


Files in this item

Thumbnail
Name:
rizk21a.pdf
Size:
1.055Mb
Format:
PDF
View/Open

This item appears in the following Collection(s)

Show simple item record