BoostingDeepNeuralNetworkEfficiencywithDual-ModuleInferenceLiuLiu12LeiDeng2ZhaodongChen2YukeWang1ShuangchenLi3JingweiZhang3YihuaYang3ZhenyuGu3YufeiDing1YuanXie2Abstract2013;Heetal.,2019;Wangetal.,2...
BeyondSyntheticNoise:DeepLearningonControlledNoisyLabelsLuJiang1DiHuang2MasonLiu3WeilongYang1AbstractHowever,duetothelackofsuitabledatasets,previousworkhasonlyexaminedDNNsoncontrolledsyntheticlabel...
BayesianSparsificationofDeepC-valuedNetworksIvanNazarov1EvgenyBurnaev1AbstractthatcausedegeneratetransformationsinR-valuednetworkswithtwicethefeaturedimensions.Theirstudydemon-Withcontinualminiatur...
AnalyticMarching:AnAnalyticMeshingSolutionfromDeepImplicitSurfaceNetworksJiabaoLei12KuiJia12Abstract(Botschetal.,2010).Forexample,oneofthemostpopu-larrepresentationsispolygonalmeshthatapproximatesa...
AnOptimisticPerspectiveonOfflineReinforcementLearningRishabhAgarwal1DaleSchuurmans12MohammadNorouzi1Abstractunsafe,orrequireahigh-fidelitysimulatorthatisoftendiffi-culttobuild(Dulac-Arnoldetal.,201...
AbstractionMechanismsPredictGeneralizationinDeepNeuralNetworksAlexGain1HavaSiegelmann2Abstractmorepuzzlingistheirabilitytomemorizerandomdatasets(Zhangetal.,2016;Yunetal.,2019).Inpartduetothis,Along...
AMean-fieldAnalysisofDeepResNetandBeyond:TowardsProvableOptimizationViaOverparameterizationFromDepthYipingLu1ChaoMa2YulongLu3JianfengLu3LexingYing4Abstract1.IntroductionTrainingDeepneuralnetworkswi...
VariationalRussianRouletteforDeepBayesianNonparametricsKaiXu1AkashSrivastava12CharlesSutton134Abstractfetprocess(IBP)priorisarepresentationlearningmethodthatinfersalatentbinaryvectorforeachdatapoin...
UnsupervisedDeepLearningbyNeighbourhoodDiscoveryJiaboHuang1QiDong1ShaogangGong1XiatianZhu2Abstractturerepresentations,whichhasrecentlydrawnincreasingattention(Wuetal.,2018;Caronetal.,2018).Deepconv...
UnderstandingandUtilizingDeepNeuralNetworksTrainedwithNoisyLabelsPengfeiChen12BenbenLiao2GuangyongChen2ShengyuZhang12AbstractbasedonwhichwecandesignspecificmethodstotrainDNNsrobustlyinpracticalappl...
TransferableClean-LabelPoisoningAttacksonDeepNeuralNetsChenZhu1W.RonnyHuang1AliShafahi1HengduoLi1GavinTaylor2ChristophStuder3TomGoldstein1Abstract2017).Incontrasttoevasionattacks(Biggioetal.,2013;S...
Trajectory-BasedOff-PolicyDeepReinforcementLearningAndreasDoerr123MichaelVolpp1MarcToussaint3SebastianTrimpe2ChristianDaniel1Abstractstandardalgorithmsarevastlydata-inefficientandrelyonmillionsofda...
TowardsAccurateModelSelectioninDeepUnsupervisedDomainAdaptationKaichaoYou12XimeiWang12MingshengLong12MichaelI.Jordan3AbstractnaturalinrecognitiontaskswhereDeepmodelshaveshowntheirsuperiority(Longet...
TowardsaDeepandUnifiedUnderstandingofDeepNeuralModelsinNLP1ChaoyuGuan2XitingWang2QuanshiZhang1RunjinChen1DiHe3XingXie2Abstract(a)Gradient-basedmethod(b)OursWedefineaunifiedinformation-basedmeasureF...
Task-AgnosticDynamicsPriorsforDeepReinforcementLearningYilunDu1KarthikNarasimhan2Abstracttt+1Whilemodel-basedDeepreinforcementlearningFigure1.Twodifferentenvironmentswithobjectdynamicsthat(RL)holds...
StochasticGradientPushforDistributedDeepLearningMahmoudAssran12NicolasLoizou13NicolasBallas1MikeRabbat1AbstracttributedtrainingofDeepnetworks(Goyaletal.,2017;Lietal.,2014).Workernodescomputelocalmi...
StochasticDeepNetworksGwendolinedeBie1GabrielPeyre´21MarcoCuturi34Abstract2012)todealnowwithshapes(Wuetal.,2015a),sounds(Leeetal.,2009),texts(LeCunetal.,1998)orMachinelearningisincreasinglytargeti...
SOLAR:DeepStructuredRepresentationsforModel-BasedReinforcementLearningMarvinZhang1SharadVikram2LauraSmith1PieterAbbeel1MatthewJ.Johnson3SergeyLevine1AbstractFigure1.Ourmethodcanlearnpoliciesforcomp...
SELFIE:RefurbishingUncleanSamplesforRobustDeepLearningHwanjunSong1MinseokKim1Jae-GilLee1Abstracttransitionmatrix(Zhangetal.,2017;Goldberger&Ben-Reuven,2017;Patrinietal.,2017;Changetal.,2017).AsOwin...
SelectiveNet:ADeepNeuralNetworkwithanIntegratedRejectOptionYonatanGeifman1RanEl-Yaniv1Abstractnetworks(DNNs).ThemostrecentrelevantpaperonthistopicisGeifman&El-Yaniv(2017),whichshowshowtoWeconsidert...