(Locally)DifferentiallyPrivateCombinatorialSemi-BanditsXiaoyuChen1KaiZheng12ZixinZhou3YunchangYang4WeiChen5LiweiWang14Abstractbasearmsandalearner(orcalledaserver)interactswiththeenvironmentforTroun...
BeatingStochasticandAdversarialSemi-BanditsOptimallyandSimultaneouslyJulianZimmert1HaipengLuo2Chen-YuWei2Abstracttrary,theminimaxoptimalregretisoforderO(√T)(Aueretal.,2002).Wedevelopthefirstgenera...
ThompsonSamplingforCombinatorialSemi-BanditsSiweiWang1WeiChen2AbstractdifferenceoverTstepsbetweenalwaysplayingthearmwiththeoptimalexpectedrewardandplayingthearmsWestudytheapplicationoftheThompsonsa...