theImpactofRecordLinkageonLearningfromFeaturePartitionedDataRichardNock1StephenHardy2WilkoHenecka2HamishIvey-Law3JakubNabaglo3GiorgioPatrini4GuillaumeSmith2BrianThorne5Abstract"whenisaglobaltrained...
theHintonsinyourNeuralNetwork:aQuantumFieldtheoryViewofDeepLearningRobertoBondesan1MaxWelling1AbstractInthispaperweproposeadirectmappingofadeepneu-ralnetworkontoaopticalquantumcomputerthroughtheInt...
theEarthMover’sPinballLoss:QuantilesforHistogram-ValuedRegressionFlorianList1AbstractFigure1.ComicofourmethodforDeepLearning-basedhistogramregression:aneuralnetwork(NN)fθwithweightsθistrainedAlt...
theEmergenceofIndividualityJiechuanJiang1ZongqingLu1Abstractmulti-cameramulti-objecttracking(Liuetal.,2017),wherecameraagentslearntocooperativelytrackmultipleobjects,Individualityisessentialinhuman...
theDistributedDiscreteGaussianMechanismforFederatedLearningwithSecureAggregationPeterKairouz1ZiyuLiu1ThomasSteinke1Abstractlearning(FL)toensurethatrawdataneverleavesusers’devices(McMahanetal.,2017...
StrategicClassificationintheDarkGaneshGhalme1VineetNair1ItayEilat1InbalTalgam-Cohen1NirRosenfeld1Abstractpayingbacktheloan.thegoalofthestrategicclassificationliteratureistoshowwhenandhowthefirmcanl...
StraighttotheGradient:LearningtoUseNovelTokensforNeuralTextGenerationXiangLin1SimengHan1ShafiqJoty12Abstractrization(Seeetal.,2017),imagecaptioning(Melas-Kyriazietal.,2018;Wang&Chan,2019)andmachine...
SparseBERT:RethinkingtheImportanceAnalysisinSelf-attentionHanShi1JiahuiGao2XiaozheRen3HangXu3XiaodanLiang4ZhenguoLi3JamesT.Kwok1AbstractincludetheBERT(Devlinetal.,2019),whichachievesstate-of-the-ar...
SKIingonSimplices:KernelInterpolationonthePermutohedralLatticeforScalableGaussianProcessesSanyamKapoor1MarcFinzi1KeAlexanderWang2AndrewGordonWilson1AbstractFigure1.Comparisonofthenumberofgridpoints...
Solvinghigh-dimensionalparabolicPDEsusingthetensortrainformatLorenzRichter123LeonSallandt4NikolasNu¨sken5AbstractonstochasticrepresentationsofthePDEsunderconsidera-tion(Eetal.,2017;Raissietal.,201...
SoftthenHard:RethinkingtheQuantizationinNeuralImageCompressionZongyuGuo1ZhizhengZhang1RunsenFeng1ZhiboChen1AbstractQuantizationisoneofthekeychallengesforneuralimagecompression.Sincethegradientofqua...
RobustDensityEstimationfromBatches:theBestThingsinLifeare(Nearly)FreeAyushJain1AlonOrlitsky1Abstractexampleshowsthatregardlessofthenumberofsamplesavailable,suchcorruptionseverelycurtailsthelearning...
RevealingtheStructureofDeepNeuralNetworksviaConvexDualityTolgaErgen1MertPilanci1Abstractwheretwo-layerReLUnetworkswiththeminimumEu-clideannormsolutionandzerotrainingerrorareprovenWestudyregularized...
RethinkingNeuralvs.Matrix-FactorizationCollaborativeFiltering:thetheoreticalPerspectivesDaXu1ChuanweiRuan2EvrenKorpeoglu1SushantKumar1KannanAchan1Abstractexploredthankstotheirinterpretabilityandcom...
RepresentationMatters:AssessingtheImportanceofSubgroupAllocationsinTrainingDataEstherRolf1theodoraWorledge1BenjaminRecht1MichaelI.Jordan12AbstractOurworkaimstodevelopageneralunifyingperspectiveonth...
QuantifyingtheBenefitofUsingDifferentiableLearningoverTangentKernelsEranMalach1PritishKamath2EmmanuelAbbe3NathanSrebro2AbstractCanallthesuccessofdeeplearningbeexplainedusingtheNTK?Thiswouldimplytha...
Puttingthe“Learning”intoLearning-AugmentedAlgorithmsforFrequencyEstimationElbertDu12FranklynWang12MichaelMitzenmacher2AbstractbyθandanalgorithmAInlearning-augmentedalgorithms,algorithmsmaxM(θ,A...
ProjectiontechniquestoupdatethetruncatedSVDofevolvingmatriceswithapplicationsVassilisKalantzis1GeorgiosKollias1ShashankaUbaru1AthanasiosN.Nikolakopoulos2LiorHoresh1KennethL.Clarkson3AbstractwhereEd...
ParetoGAN:ExtendingtheRepresentationalPowerofGANstoHeavy-TailedDistributionsToddHuster1JeremyE.J.Cohen1ZinanLin2KevinChan3CharlesKamhoua3NandiLeslie4Cho-YuJasonChiang1VyasSekar2AbstractEvans&Erland...
OutsidetheEchoChamber:OptimizingthePerformativeRiskJohnMiller1JuanC.Perdomo1TijanaZrnic1Abstractriumnotionknownasperformativestability(Drusvyatskiy&Xiao,2020;Mendler-Du¨nneretal.,2020;Brownetal.,I...