PrivateAdaptiveGradientMethodsforConvexOptimizationHilalAsi12JohnDuchi23AlirezaFallah41OmidJavidbakht5KunalTalwar5Abstractopingprivatevariantsofstochasticgradientdescent(SGD),wherealgorithmsguarant...
PrincipledSimplicialNeuralNetworksforTrajectoryPredictionT.MitchellRoddenberry1NicholasGlaze1SantiagoSegarra1Abstractintheirabilitytoincorporatearbitrarypairwiserelationalstructuresintheircomputati...
PrincipalComponentHierarchyforSparseQuadraticProgramsRobbieVreugdenhil1VietAnhNguyen23ArminEftekhari4PeymanMohajerinEsfahani1Abstractspecifically,weconsidertheproblemWeproposeanovelapproximationhie...
PosteriorValueFunctions:HindsightBaselinesforPolicyGradientMethodsChrisNota1BrunoCastrodaSilva1PhilipS.Thomas1Abstractcases,suchinformationcanbeusefulforassessingwhichoutcomeswerelikelytohaveoccurr...
PopSkipJump:Decision-BasedAttackforProbabilisticClassifiersCarl-JohannSimon-Gabriel1NomanAhmedSheikh1AndreasKrause1Abstractnoisyorprobabilisticclassificationoutputs–aquitenaturalandcommonsettingin...
PolicyInformationCapacity:Information-TheoreticMeasureforTaskComplexityinDeepReinforcementLearningHirokiFuruta1TatsuyaMatsushima1TadashiKozuno2YutakaMatsuo1SergeyLevine3OfirNachum3ShixiangShaneGu3A...
PolicyGradientBayesianRobustOptimizationforImitationLearningZaynahJaved1DanielS.Brown1SatvikSharma1JerryZhu1AshwinBalakrishna1MarekPetrik2AncaD.Dragan1KenGoldberg1Abstracthuman-designedrewardfuncti...
PipeTransformer:AutomatedElasticPipeliningforDistributedTrainingofLarge-scaleModelsChaoyangHe1ShenLi2MahdiSoltanolkotabi1SalmanAvestimehr1AbstractTransformer(ViT)(Dosovitskiyetal.,2020)alsoachieved...
ParametricGraphforUnimodalRankingBanditCamille-SovannearyGauthier12RomaricGaudel3ElisaFromont452BoammaniAserLompo6Abstractuserattention.Typicalexamplesofsuchdisplaysare(i)alistofnews,visibleonebyon...
ParameterlessTransductiveFeatureRe-representationforFew-ShotLearningWentaoCui1YuhongGuo12AbstractFSLsolutionsaregenerallydevelopedintwobranches:meta-learningbasedmethodsandfinetuningbasedmeth-Recen...
PAC-LearningforStrategicClassificationRaviSundaram1AnilVullikanti23HaifengXu2FanYao2AbstracttheongoingCOVID-19pandemic(Bryan&Crossroads;Williams&Haire).Intheearlymonthsofthepandemic,Thestudyofstrat...
Order-AgnosticCrossEntropyforNon-AutoregressiveMachineTranslationCunxiaoDu1ZhaopengTu2JingJiang1AbstractAnumberofrecenteffortshaveexploredwaystoimprovetheNATmodels’abilitytohandlemultimodality.One...
OptimizationPlanningfor3DConvNetsZhaofanQiu1TingYao1Chong-WahNgo2TaoMei1Abstractstance,anensembleofLGD-3Dnetworks(Qiuetal.,2019)achieves17.88%intermsofaverageerrorintrimmedvideoItisnottrivialtoopti...
OptimalTransportKernelsforSequentialandParallelNeuralArchitectureSearchVuNguyen∗1TamLe∗2MakotoYamada23MichaelA.Osborne4Abstractreaderstothesurvey(Elskenetal.,2019b)foradetailedreviewofNASandtothe...
OptimalThompsonSamplingstrategiesforsupport-awareCVaRbanditsDorianBaudry1RomainGautron23EmilieKaufmann1Odalric-AmbrymMaillard1AbstractValueatRisk(CVaR)aswellasmoregenericcoherentspec-tralriskmeasur...
OptimalStreamingAlgorithmsforMulti-ArmedBanditsTianyuanJin1KekeHuang1JingTang2XiaokuiXiao1Abstractson,1933),onlineadvertisement(Bertsimas&Mersereau,2007),andcrowdsourcing(Zhouetal.,2014).Ittypicall...
OptimalregretalgorithmforPseudo-1dBanditConvexOptimizationAadirupaSaha1NagarajanNatarajan2PraneethNetrapalli23PrateekJain23Abstracttheproblemhasa"pseudo-1d"structureinthelossfunc-tionsft(w)=t(gt(w;...
OopsITookAGradient:ScalableSamplingforDiscreteDistributionsWillGrathwohl12KevinSwersky2MiladHashemi2DavidDuvenaud12ChrisJ.Maddison1AbstractFigure1.Ourapproachvisualized.Oftendiscretedistributionsar...
On-the-FlyRectificationforRobustLarge-VocabularyTopicInferenceMoontaeLee1SungjunCho2KunDong3DavidMimno4DavidBindel5Abstractbetweenthelatentvariables(Bleietal.,2003;Airoldietal.,2008;A.Erosheva,2003...
On-OffCenter-SurroundReceptiveFieldsforAccurateandRobustImageClassificationZahraBabaiee1RaminHasani2MathiasLechner3DanielaRus2RaduGrosu1AbstractOOCS-CNNFeedforwardRobustnesstovariationsinlightingco...