Tiltingtheplayingfield:DynamicallossFunctionsformachinelearningMiguelRuiz-Garcia12GeZhang1SamuelS.Schoenholz3AndreaJ.Liu1Abstracttionscale(Glorot&Bengio,2010;Xiaoetal.,2018)orlearningrateschedule(H...
PosteriorValueFunctions:HindsightBaselinesforPolicyGradientMethodsChrisNota1BrunoCastrodaSilva1PhilipS.Thomas1Abstractcases,suchinformationcanbeusefulforassessingwhichoutcomeswerelikelytohaveoccurr...
PhaseTransitions,DistanceFunctions,andImplicitNeuralRepresentationsYaronLipman12AbstractFigure1.ImplicitneuralrepresentationslearnedwiththenovelPHASEloss.Thisnewlossunifiesoccupancyandsigneddistanc...
OptimizingpersistenthomologybasedFunctionsMathieuCarrie`re1Fre´de´ricChazal2MarcGlisse2YuichiIke3HariprasadKannan2YuheiUmeda3Abstractthelastfewyears,agrowingnumberofworkshavesuc-cessfullybeenusin...
MediatedUncoupledLearning:LearningFunctionswithoutDirectInput-outputCorrespondencesIkkoYamane12JunyaHonda32FlorianYger12MasashiSugiyama24Abstracttionpredictingsentimentofanimage(Mittaletal.,2018).E...
AsymmetricLossFunctionsforLearningwithNoisyLabelsXiongZhou12XianmingLiu12JunjunJiang12XinGao32XiangyangJi4AbstractToalleviatetheimpactoflabelnoisetoclassifierlearning,onepopularresearchlineistodesi...
AFunctionalPerspectiveonLearningSymmetricFunctionswithNeuralNetworksAaronZweig1JoanBruna12AbstractSuchsymmetricFunctionsappearnaturallyacrossseveraldomains,includingparticlephysics,computergraphics...
Super-efficiencyofautomaticdifferentiationforFunctionsdefinedasaminimumPierreAblin1GabrielPeyré1ThomasMoreau2Abstractsparsecode,andListheLassocost(Mairaletal.,2010).Inthiscase,measurestheabilityof...
StochasticRank:GlobalOptimizationofScale-FreeDiscreteFunctionsAlekseiUstimenko1LiudmilaProkhorenkova123AbstractTodealwiththediscretestructureofarankingloss,onecanusesomesmoothapproximation,whichise...
PeerLossFunctions:LearningfromNoisyLabelswithoutKnowingNoiseRatesYangLiu1HongyiGuo2Abstracttensivepriorworksintheliteraturethataimtodevelopal-gorithmstolearnmodelsthatarerobusttolabelnoise(By-Learn...
OptimizationfromStructuredSamplesforCoverageFunctionsWeiChen1XiaomingSun23JialinZhang23ZhijieZhang23Abstracteters.Thecomputationaltaskistooptimizeanobjectivefunctiongivenparametersofthemodel.Onesuc...
OnSecond-OrderGroupInfluenceFunctionsforBlack-BoxPredictionsSamyadeepBasu1XuchenYou1SoheilFeizi1Abstracttotracebackmodelpredictionstoitstrainingdata.Morespecifically,onecanaskwhichtrainingsampleswe...
NormalizedLossFunctionsforDeepLearningwithNoisyLabelsXingjunMa1HanxunHuang1YisenWang2SimoneRomanoSarahErfani1JamesBailey1Abstract2018;Hanetal.,2018b);and4)robustlossFunctionsthatareinherentlytolera...
InterpretingRobustOptimizationviaAdversarialInfluenceFunctionsZhunDeng1CynthiaDwork1JialiangWang1LinjunZhang2AbstracttionforamodelMisformulatedasRobustoptimizationhasbeenwidelyusedinminEx,ymaxl(θM...
FastandPrivateSubmodularandk-SubmodularFunctionsMaximizationwithMatroidConstraintsAkbarRafiey1YuichiYoshida2AbstractcombinatorialoptimizationproblemsincludingtheMaxi-mumCoverage,MaximumCut,andFacil...
EfficientlySamplingFunctionsfromGaussianProcessPosteriorsJamesT.Wilson1ViacheslavBorovitskiy23AlexanderTerenin1PeterMostowsky2MarcPeterDeisenroth4Abstracttionandlong-termversusshort-termrewards.Bay...
ComposableSketchesforFunctionsofFrequencies:BeyondtheWorstCaseEdithCohen12OfirGeri34RasmusPagh156Abstractdatawithsmallcomputationalresources(time,communi-cation,andspace).Suchsketchessupportprocess...
ValidatingCausalInferenceModelsviaInfluenceFunctionsAhmedM.Alaa1MihaelavanderSchaar123AbstractFigure1.Pictorialrepresentationofourvalidationprocedure.Toestimatetheperformancesoftwocompetingmodels(m...