ThePriceofDifferentialPrivacyforOnlineLearningNamanAgarwal1KaranSingh1Abstractthefullinformationandpartialinformation(bandit)set-tings.ThisresultimprovestheknownbestregretboundsWedesigndifferential...
Tensor-TrainRecurrentNeuralNetworksforVideoClassificationYinchongYang12DenisKrompass2VolkerTresp12AbstractralNetworkarchitectures.ThissuccessmotivateseffortstoalsoapplytheseRNNstovideodata,sinceavi...
Sub-sampledCubicRegularizationforNon-convexOptimizationJonasMoritzKohler1AurelienLucchi1Abstractlearningwhennisverylarge.Itsconvergenceproper-tiesarewellunderstoodforconvexfunctions,whichariseWecon...
StochasticVarianceReductionMethodsforPolicyEvaluationSimonS.Du1JianshuChen2LihongLi2LinXiao2DengyongZhou2Abstractimportantinformationfortheagenttooptimizeitspolicy.forexample,policy-iterationalgori...
StrongNP-HardnessforSparseOptimizationwithConcavePenaltyFunctionsYichenChen1DongdongGe2MengdiWang1ZizhuoWang3YinyuYe4HaoYin4AbstractWeareinterestedinthecomputationalcomplexityofProb-lem1undergenera...
StochasticGradientMCMCMethodsforHiddenMarkovModelsYi-AnMa1NicholasJ.Foti1EmilyB.Fox1Abstractdistributioninthepresenceofsuchnoise.Significanthead-wayhasbeenmadeindevelopingsuchcorrectSG-MCMCStochast...
StochasticDCAfortheLarge-sumofNon-convexFunctionsProblemanditsApplicationtoGroupVariableSelectioninClassificationHoaiAnLeThi1HoaiMinhLe1DuyNhatPhan1BachTran1AbstractNowadays,thegrowthoftechnologies...
StochasticAdaptiveQuasi-NewtonMethodsforMinimizingExpectedValuesChaoxuZhou∗1WenboGao∗1DonaldGoldfarb1AbstractinfeasibletoanalyticallycomputeEf(x,ξ).ThiscanberesolvedbyreplacingtheexpectationEf(x...
StatisticalInferenceforIncompleteRankingData:TheCaseofRank-DependentCoarseningMohsenAhmadiFahandar1EykeHu¨llermeier1Ine´sCouso2Abstract“items”),whereasnopreferencesarerevealedaboutthere-maining...
StabilisingExperienceReplayforDeepMulti-AgentReinforcementLearningJakobFoerster1NantasNardelli1GregoryFarquhar1TriantafyllosAfouras1Philip.H.S.Torr1PushmeetKohli2ShimonWhiteson1Abstractmulti-agents...
SphericalStructuredFeatureMapsforKernelApproximationYuemingLyu1Abstractdenotesthenonlinearfunction,W∈Rd×NisconstructedbyNi.i.dsamplesdrawnfromadistributiondefinedbyWeproposeSphericalStructuredFea...
Source-TargetSimilarityModelingsforMulti-SourceTransferGaussianProcessRegressionPengfeiWei12RamonSagarna12YipingKe12Yew-SoonOng12Chi-KeongGoh32Abstracttion(Tommasietal.,2014;Fangetal.,2015;Bhatteta...
Soft-DTW:aDifferentiableLossFunctionforTime-SeriesMarcoCuturi1MathieuBlondel2AbstractInputOutputWeproposeinthispaperadifferentiablelearningFigure1.Giventhefirstpartofatimeseries,wetrainedtwolossbet...
SlicedWassersteinKernelforPersistenceDiagramsMathieuCarrie`re1MarcoCuturi2SteveOudot1Abstractafew.ThetoolsdevelopedinTDAarebuiltuponpersis-tenthomologytheory(Edelsbrunner&Harer,2010;Oudot,Persisten...
SimultaneousLearningofTreesandRepresentationsforExtremeClassificationandDensityEstimationYacineJernite1AnnaChoromanska1DavidSontag2AbstractHinton,2009),extrinsichierarchies(Morin&Bengio,2005)whichc...
SharpMinimaCanGeneralizeforDeepNetsLaurentDinh1RazvanPascanu2SamyBengio3YoshuaBengio14Abstractapproximatecertainfunctions(e.g.Montufaretal.,2014;Raghuetal.,2016).Otherworks(e.gDauphinetal.,2014;Des...
SelectiveInferenceforSparseHigh-OrderInteractionModelsShinyaSuzumura1KazuyaNakagawa1YutaUmezu1KojiTsuda23IchiroTakeuchi13AbstractFigure1.Exampleofthetreestructureamonghigh-orderinterac-tionfeatures...
ScalableGenerativeModelsforMulti-labelLearningwithMissingLabelsVikasJain1NirbhayModhe1PiyushRai1Abstracttisingandrecommendersystems(Prabhu&Varma,2014;Jainetal.,2016),etc.Wepresentascalable,generati...
SARAH:ANovelMethodforMachineLearningProblemsUsingStochasticRecursiveGradientLamM.Nguyen1JieLiu1KatyaScheinberg12MartinTaka´cˇ1AbstractProblemsofthistypearisefrequentlyinsupervisedlearn-Inthispape...
Safety-AwareAlgorithmsforAdversarialContextualBandit122WenSunDebadeeptaDeyAshishKapoorAbstractside-effectofanewtreatmentmustbetakenintoconsidera-tionforpatients’safety.Ingeneraltheseapplicationswi...