ProvablyStrictGeneralisationBenefitforEquivariantModelsBrynElesedy1SheheryarZaidi2AbstractasSokolicetal.(2017);Sannai&Imaizumi(2019),coveronlytheworstcaseperformanceofalgorithms.TheseworksItiswidel...
ProvablyEnd-to-endLabel-noiseLearningwithoutAnchorPointsXuefengLi12TongliangLiu2BoHan3GangNiu4MasashiSugiyama45Abstract(Arpitetal.,2017;Zhangetal.,2017;Xiaetal.,2021;Wuetal.,2021).Inlabel-noiselear...
ProvablyEfficientReinforcementLearningforDiscountedMDPswithFeatureMappingDongruoZhou1JiafanHe1QuanquanGu1Abstractlinearfunctionsorneuralnetworkstomapstatesandactionstoalow-dimensionalspaceandsolvet...
ProvablyEfficientLearningofTransferableRewardsAlbertoMariaMetelli1GiorgiaRamponi1AlessandroConcetti1MarcelloRestelli1Abstracttheoretically,underthestrongassumptionofrewardunique-ness(Abbeel&Ng,2004...
ProvablyEfficientFictitiousPlayPolicyOptimizationforZero-SumMarkovGameswithStructuredTransitionsShuangQiu1XiaohanWei2JiepingYe1ZhaoranWang3ZhuoranYang4Abstractunderstandingofmulti-agentpolicyoptimi...
ProvablyEfficientAlgorithmsforMulti-ObjectiveCompetitiveRLTianchengYu1YiTian1JingzhaoZhang1SuvritSra1Abstractaveragereturntoatargetsetsmallaslongasthissetsatisfiesaconditioncalledapproachability(Bl...
ProvablyCorrectOptimizationandExplorationwithNon-linearPoliciesFeiFeng1WotaoYin1AlekhAgarwal2LinYang3Abstractrer&Geist,2014;Geistetal.,2019;Abbasi-Yadkorietal.,2019;Agarwaletal.,2020c;Bhandari&Russ...
IsPessimismProvablyEfficientforOfflineRL?YingJin1ZhuoranYang2ZhaoranWang3AbstractVinyalsetal.,2017)reliesontwoingredients:(i)expressivefunctionapproximators,e.g.,deepneuralnetworks(LeCunWestudyoffl...
ProvablyEfficientModel-basedPolicyAdaptationYudaSong1AditiMavalankar1WenSun2SicunGao1AbstractMordatchetal.,2015),ormeta-learnpoliciesormodelsthatcanbequicklyadaptedtoin-distributionenvironments(Fin...
ProvablyEfficientExplorationinPolicyOptimizationQiCai1ZhuoranYang2ChiJin3ZhaoranWang1Abstractofiterations,evengiveninfinitedata.Meanwhile,fromthestatisticalperspective,itremainsunclearhowtoattainWh...
ProvablyConvergentTwo-TimescaleOff-PolicyActor-CriticwithFunctionApproximationShangtongZhang1BoLiu2HengshuaiYao3ShimonWhiteson1Abstractatwo-timescaleconvergentanalysisunderfunctionapproxi-mation(Ko...
KinematicStateAbstractionandProvablyEfficientRich-ObservationReinforcementLearningDipendraMisra1MikaelHenaff1AkshayKrishnamurthy1JohnLangford1Abstractfromthewell-studiedtabularsettingtoexploretheen...
ProvablyefficientRLwithRichObservationsviaLatentStateDecodingSimonS.Du1AkshayKrishnamurthy2NanJiang3AlekhAgarwal4MiroslavDud´ık2JohnLangford2Abstract2010;Lattimore&Hutter,2012).Consequently,treat...
ProvablyEfficientImitationLearningfromObservationAloneWenSun1AnirudhVemula1ByronBoots2J.AndrewBagnell3Abstractaction,viasupervisedlearningapproaches(e.g.,DAgger(Rossetal.,2011),AggreVaTe(Ross&Bagne...
ProvablyEfficientMaximumEntropyExplorationEladHazan12ShamM.Kakade342KaranSingh12AbbyVanSoest12Abstractsuchaslearningwithintrinsicrewardandcuriositydrivenmethods,surveyedbelow.Ourworkstudiesaclassof...
Plug-and-PlayMethodsProvablyConvergewithProperlyTrainedDenoisersErnestK.Ryu1JialinLiu1SichengWang2XiaohanChen2ZhangyangWang2WotaoYin1Abstractmeasurementsoftheimage,isencodedinf(x).Sof(x)issmallifxa...
WidthProvablyMattersinOptimizationforDeepLinearNeuralNetworksSimonS.Du1WeiHu2Abstractconvergestoglobalminimumunderfurtherassumptionsonbothdataandglobalminimum.TheseresultsrequireWeprovethatforanL-l...
Quickshift++:ProvablyGoodInitializationsforSample-BasedMeanShiftHeinrichJiang1JenniferJang2SamoryKpotufe3AbstractOneofthedrawbacksofthesetwoprocedures,aswellasmanymode-seekingbasedclusteringalgorit...
DifferentiableAbstractInterpretationforProvablyRobustNeuralNetworksMatthewMirman1TimonGehr1MartinVechev1Abstractsarialattackswhichprovidescertificatesprovingthatnoneofthetrainingexamplescouldbeadve...
ProvablyOptimalAlgorithmsforGeneralizedLinearContextualBanditsLihongLi1YuLu2DengyongZhou1Abstractetal.,2009;Lietal.,2010;2012).Intheproblemofper-sonalizednewsrecommendation,thewebsitemustrecom-Cont...
