GoodSubnetworksProvablyExist:PruningviaGreedyForwardSelectionMaoYe1ChengyueGong1LizhenNie2DennyZhou3AdamKlivans1QiangLiu1Abstractachievinghuman-levelaccuracyoncomplexcognitivetasks,includingimagecl...
LearningHumanObjectivesbyEvaluatingHypotheticalBehaviorSiddharthReddy1AncaD.Dragan1SergeyLevine1ShaneLegg2JanLeike2AbstractGenerativemodelHypotheticalbehaviorWeseektoalignagentbehaviorwithauser’so...
LearningDiscreteStructuredRepresentationsbyAdversariallyMaximizingMutualInformationKarlStratos1SamWiseman2Abstractoverlatentvariableswithoutmodelingrawsignalsbymax-imizingthemutualinformationbetwee...
InterpretableOff-PolicyEvaluationinReinforcementLearningbyHighlightingInfluentialTransitionsOmerGottesman1JosephFutoma1YaoLiu2SonaliParbhoo1LeoAnthonyCeli3EmmaBrunskill2FinaleDoshi-Velez1Abstractan...
InductiveRelationPredictionbySubgraphReasoningKomalK.Teru12EtienneG.Denis12WilliamL.Hamilton12AbstractFigure1.Illustrationoftransductiveandinductivesettingsforre-lationpredictioninknowledgegraphs.T...
ImprovingMolecularDesignbyStochasticIterativeTargetAugmentationKevinYang1WengongJin2KyleSwanson3ReginaBarzilay2TommiJaakkola2Abstractascomplexgraph-structuredobjects.Asaresult,thesemod-elsrequireco...
ImprovingGeneralizationbyControllingLabel-NoiseInformationinNeuralNetworkWeightsHrayrHarutyunyan1KyleReing1GregVerSteeg1AramGalstyan1Abstractproposednoregularizationdropout0.5weightdecay0.003Inthep...
IdentifyingRewardFunctionsusingAnchorActionsSinongGeng1HoussamNassif2CarlosA.Manzanares2A.MaxReppen3RonnieSircar3Abstractwithfirmprofitfunctions(Abbring,2010;AguirregabiriaandNevo,2013).Weproposear...
Few-shotDomainAdaptationbyCausalMechanismTransferTakeshiTeshima12IsseiSato12MasashiSugiyama21AbstractFigure1:Nonparametricgen-erativemodelofnonlinearin-Westudyfew-shotsuperviseddomainadaptationdepe...
FamilywiseErrorRateControlbyInteractiveUnmaskingBoyanDuan1AadityaRamdas12LarryWasserman12Abstractwhenµ<0,ithasnondecreasingdensity.Alowp-valuesuggestsevidencetorejectthenullhypothesis.Weproposeame...
EnhancingSimpleModelsbyExploitingWhatTheyAlreadyKnowAmitDhurandhar1KarthikeyanShanmugam1RonnyLuss1Abstract1.IntroductionTherehasbeenrecentinterestinimprovingper-Simplemodelssuchasdecisiontreesorrul...
Communication-EfficientDistributedPCAbyRiemannianOptimizationLong-KaiHuang1SinnoJialinPan1Abstractfollowingempiricalriskminimization(ERM)problemInthispaper,westudytheleadingeigenvec-w∗=argminFˆ(w...
BoostingFrank-WolfebyChasingGradientsCyrilleW.Combettes1SebastianPokutta23AbstractifC={X∈Rm×nXnucτ}isanuclearnorm-ball,aprojectionontoCrequirescomputinganSVD,whichhasTheFrank-Wolfealgorithmhasbe...
BayesianExperimentalDesignforImplicitModelsbyMutualInformationNeuralEstimationStevenKleinegesse1MichaelU.Gutmann1Abstractology(Coranderetal.,2017),cellbiology(Voetal.,2015)andpharmacokinetics(Donne...
AmortisedLearningbyWake-SleepLiK.Wenliang1TheodoreMoskovitz1HeishiroKanagawa1ManeeshSahani1Abstractprominentandcomeswithattractivetheoreticalproperties,includingconsistencyandasymptoticefficiency(N...
Acceleratingthediffusion-basedensemblesamplingbynon-reversibledynamicsFutoshiFutami12IsseiSato12MasashiSugiyama21AbstractMurphy,2012).Formanypracticalmodels,wecannotob-tainananalyticalexpressionoft...
UnsupervisedDeepLearningbyNeighbourhoodDiscoveryJiaboHuang1QiDong1ShaogangGong1XiatianZhu2Abstractturerepresentations,whichhasrecentlydrawnincreasingattention(Wuetal.,2018;Caronetal.,2018).Deepconv...
State-ReificationNetworks:ImprovingGeneralizationbyModelingtheDistributiAonoAfHiddenR10.e0BpreBsentations10.010.0xx10.07.57.55.05.05.05.02.52.50.00.00.00.0AlexLamb1JonathanBinas1AnirudhGoyal2.52.51...
Noise2Self:BlindDenoisingbySelf-SupervisionJoshuaBatson1LoicRoyer1Abstracthighlycorrelated.Speakingloosely,ifthe“latentdimen-sion”ofthespaceofobjectsunderstudyismuchlowerthanWeproposeageneralfram...
MONK–Outlier-RobustMeanEmbeddingEstimationbyMedian-of-MeansMatthieuLerasle12Zolta´nSzabo´3Timothe´eMathieu1GuillaumeLecue´4Abstractversatilityofdomainswherekernelsexist;examplesincludetrees(Co...
