RobustAsymmetricLearninginPOMDPsAndrewWarrington1J.WilderLavington23AdamS´cibior23MarkSchmidt24FrankWood235Abstracttheworld,tocompletethetask.Atrainee,observingonlyimages,canthenlearntomimictheact...
ProjectionRobustWassersteinBarycentersMinhuiHuang1ShiqianMa2LifengLai1AbstractHowever,computingtheWBforasetofprobabilitydis-tributionsisnotoriouslyhard.ThehardnesscomesfromCollectingandaggregatingi...
PolicyGradientBayesianRobustOptimizationforImitationLearningZaynahJaved1DanielS.Brown1SatvikSharma1JerryZhu1AshwinBalakrishna1MarekPetrik2AncaD.Dragan1KenGoldberg1Abstracthuman-designedrewardfuncti...
On-the-FlyRectificationforRobustLarge-VocabularyTopicInferenceMoontaeLee1SungjunCho2KunDong3DavidMimno4DavidBindel5Abstractbetweenthelatentvariables(Bleietal.,2003;Airoldietal.,2008;A.Erosheva,2003...
OnRobustMeanEstimationunderCoordinate-levelCorruptionZifanLiu1JonghoPark1TheodorosRekatsinas1ChristosTzamos1Abstractfilteringordown-weightingcorrupteddatavectorstoreducetheirinfluence(Diakonikolase...
OnLowerBoundsforStandardandRobustGaussianProcessBanditOptimizationXuCai1JonathanScarlett12Abstractgunovicetal.,2016;Wang&Jegelka,2017;Janzetal.,2020),andalgorithm-independentlowerboundshavebeenInth...
MonotonicRobustPolicyOptimizationwithModelDiscrepancyYuankunJiang1ChenglinLi2WenruiDai1JunniZou1HongkaiXiong2Abstractcontroltasks,e.g.,playingcomputergameswithhuman-levelperformance(Mnihetal.,2013;...
MakingPaperReviewingRobusttoBidManipulationAttacksRuihanWu1ChuanGuo2FelixWu3†RahulKidambi4†LaurensvanderMaaten2KilianQ.Weinberger1Abstractbidsisimportantbecausethereviewqualityishigherwhenreviewe...
MakingtransportmoreRobustandinterpretablebymovingdatathroughasmallnumberofanchorpointsChi-HengLin1MehdiAzabou12EvaL.Dyer123Abstracttivemodeling(MartinArjovsky&Bottou,2017;Tolstikhinetal.,2017),docu...
LearningfromHistoryforByzantineRobustOptimizationSaiPraneethKarimireddy1LieHe1MartinJaggi1Abstracttryingtoderailtheprocess,ormightsimplybemalfunc-tioningandhencesendingarbitrarymessages.EnsuringByz...
ImprovedCorruptionRobustAlgorithmsforEpisodicReinforcementLearningYifangChen1SimonS.Du1KevinJamieson1Abstractstageaccordingtotheunderlyingtransitionfunction.Westudyepisodicreinforcementlearningunde...
First-OrderMethodsforWassersteinDistributionallyRobustMDPsJulienGrand-Cle´ment1ChristianKroer1Abstractpolicies,astheyoptimizeonlyfortheworst-casekernelre-alization,withoutincorporatingdistribution...
Expressive1-LipschitzNeuralNetworksforRobustMultipleGraphLearningagainstAdversarialAttacksXinZhao1ZeruZhang1ZijieZhang1LingfeiWu2JiayinJin1YangZhou1RuomingJin3DejingDou45DaYan6Abstractment)(Zhang&T...
EfficientTrainingofRobustDecisionTreesAgainstAdversarialExamplesDanie¨lVos1SiccoVerwer1Abstractetal.,2019),wecloselymimicthegreedyrecursivesplit-tingstrategythattraditionaldecisiontreesuseandwesco...
SPECTRE:DefendingAgainstBackdoorAttacksUsingRobustStatisticsJonathanHayase1WeihaoKong1RaghavSomani1SewoongOh1AbstractaccuracyonpoisonedtestexamplesStartingwiththeseminalworkof(Guetal.,2017),thereha...
CumulantsofHawkesProcessesareRobusttoObservationNoiseWilliamTrouleau1JalalEtesami2MatthiasGrossglauser1NegarKiyavash2PatrickThiran1Abstracttheyareusedtomodelthestochastictimeevolutionoflimitorderbo...
CRFL:CertifiablyRobustFederatedLearningagainstBackdoorAttacksChulinXie1MinghaoChen2Pin-YuChen3BoLi1AbstractCRFLTrainingCRFLTestingFederatedLearning(FL)asadistributedlearn-ModelUpdatesParameteringpa...
DualPrincipalComponentPursuitforRobustSubspaceLearning:TheoryandAlgorithmsforaHolisticApproachTianyuDing1ZhihuiZhu2Rene´Vidal3DanielP.Robinson4Abstractoutliersandnoise.Unlikeinliers,whichexactlyli...
DoublyRobustOff-PolicyActor-Critic:ConvergenceandOptimalityTengyuXu1ZhuoranYang2ZhaoranWang3YingbinLiang1Abstract(Haarnojaetal.,2018),etc.However,thesesuccessesusu-allyrelyontheaccesstoon-policysam...
DORO:DistributionalandOutlierRobustOptimizationRuntianZhai1ChenDan1J.ZicoKolter1PradeepRavikumar1Abstractshiftproblems,suchaslearningforalgorithmicfairness(Dworketal.,2012;Barocas&Selbst,2016)where...