onMediansof(Randomized)PairwiseMeansPierreLaforgue1StephanCle´menc¸on1PatriceBertail2Abstract1.IntroductionTournamentprocedures,recentlyintroducedinInLugosi&Mendelson(2016),theconceptoftournament...
onEfficientOptimalTransport:AnAnalysisofGreedyandAcceleratedMirrorDescentAlgorithmsTianyiLin1NhatHo2MichaelI.Jordan3Abstractscalableposteriorsamplingforlargedatasets(Srivastavaetal.,2015;2018),opti...
onLearningInvariantRepresentationsforDomainAdaptationHanZhao1RemiTachetdesCombes2KunZhang1GeoffreyJ.Gordon12Abstractcomputationalpower.However,inmanycriticalapplica-tions,e.g.,self-drivingcarsorper...
onDiscriminativeLearningofPredictionUncertaintyVojtechFranc1DanielPrusa1Abstractplug-inruleobtainedafterreplacingtheposteriorprobabili-tiesbytheirestimates.Aqualityoftheplug-inruleheavilyInclassifi...
onDropoutandNuclearNormRegularizationPooryaMianjy1RamanArora1Abstractislinear,therepresentationinfactoredformmakestheop-timizationlandscapenon-convexandhence,challengingWegiveaformalandcompletechar...
onConnectedSublevelSetsinDeepLearningQuynhNguyen1Abstractcontinuouspathfromanystartingpointinparameterspaceonwhichthelossisnon-increasingandgetsarbitrarilycloseThispapershowsthateverysublevelsetoft...
onCertifyingNon-uniformBoundsagainstAdversarialAttacksChenLiu1RyotaTomioka2VolkanCevher1AbstractCWattack(Carlini&Wagner,2017).MostattackmethodssearchfortheadversarialexamplebyutilizingthegradientTh...
NewResultsonInformationTheoreticClusteringFerdinandoCicalese1EduardoLaber2LucasMurtinho2Abstracttropy)thatestimatethedissimilarityofagroupofitems(see,e.g.,(Dhillonetal.,2003)andreferencestherein)In...
First-OrderAlgorithmsConvergeFasterthanO(1/k)onConvexProblemsChing-peiLee1StephenJ.Wright1Abstract(3)reliesonshowingthatItiswellknownthatbothgradientdescentandkstochasticcoordinatedescentachieveagl...
DoublyRobustJointLearningforRecommendationonDataMissingNotatRandomXiaojieWang1RuiZhang1YuSun2JianzhongQi1Abstract(MNAR).Forexample,arecentstudyinsongrecommen-dationshowsthattheprobabilityofaratingb...
DataPoisoningAttacksonStochasticBanditsFangLiu1NessShroff12Abstractismotivatedbymodernindustrialscaleapplicationsofma-chinelearningsystems,wheredatacollectionandpolicyStochasticmulti-armedbanditsfo...
AnOptimalPrivateStochastic-MABAlgorithmBasedonanOptimalPrivateStoppingRuleTouqirSajed1OrSheffet1Abstractexperimentsinmedicine(Robbins,1952),hasapplicationsinfieldssuchasranking(Kvetonetal.,2015),re...
Adversarialcamerastickers:Aphysicalcamera-basedattackondeeplearningsystemsJunchengB.Li12FrankR.Schmidt1J.ZicoKolter12AbstractFigure1.Illustrationofourapproach:(left)ouradversarialstickeraffixedtoac...
AdversarialAttacksonNodeEmbeddingsviaGraphPoisoningAleksandarBojchevski1StephanGu¨nnemann1Abstract&Wagner,2017;Liangetal.,2018;Cisse´etal.,2017;Linetal.,2017;Chenetal.,2017a).Thegoalofnetworkrepr...
AWrappedNormalDistributiononHyperbolicSpaceforGradient-BasedLearningYoshihironagano1ShoichiroYamaguchi2YasuhiroFujita2MasanoriKoyama2Abstract(a)AtreerepresentationofthetrainingdatasetHyperbolicspac...
ALarge-ScaleStudyonRegularizationandNormalizationinGANsKarolKurach1MarioLucic1XiaohuaZhai1MarcinMichalski1SylvainGelly1Abstractfromthetruedistributionorweresynthesizedbythegenera-tor.Thesolutiontot...
ADynamicalSystemsPerspectiveonNesterovAccelerationMichaelMuehlebach1MichaelI.Jordan1Abstractbeenmanyattemptstounderstandandcharacterizethephe-nomenon.Bubecketal.(2015)suggestamodificationofWepresen...
ADeepReinforcementLearningPerspectiveonInternetCongestionControlNathanJay1NogaH.Rotman2P.BrightenGodfrey1MichaelSchapira2AvivTamar3AbstractFigure1:MultipletrafficflowssharingalinkWepresentandinvest...
Semi-SupervisedLearningonDataStreamsviaTemporalLabelPropagationTalWagner1SudiptoGuha2ShivaPrasadKasiviswanathan2NinaMishra2Abstractingofmetricsarisingfrommedicalpatientsignals(ECG,EEG,falldetection...
RepresentationLearningonGraphswithJumpingKnowledgeNetworksKeyuluXu1ChengtaoLi1YonglongTian1TomohiroSonobe2Ken-ichiKawarabayashi2StefanieJegelka1Abstract2017;Grover&Leskovec,2016;Tangetal.,2015).Rec...