Non-NegativeBregmanDivergenceMinimizationforDeepDirectDensityRatioEstimationMasahiroKato1TakeshiTeshima2Abstractthedensityratiowhenanomaly-freeunlabeledtestdataareavailable(Hidoetal.,2008).Densityr...
MediatedUncoupledLearning:LearningFunctionswithoutDirectInput-outputCorrespondencesIkkoYamane12JunyaHonda32FlorianYger12MasashiSugiyama24Abstracttionpredictingsentimentofanimage(Mittaletal.,2018).E...
LearningRandomlyPerturbedStructuredPredictorsforDirectLossMinimizationHeddaCohenIndelman1TamirHazan1Abstractgradientsoftwopredictedstructures:overtheoriginallossfunctionandoveraperturbedlossfunctio...
Memory-OptimalDirectConvolutionsforMaximizingClassificationAccuracyinEmbeddedApplicationsAlbertGural1BorisMurmann1Abstract(IoT),themachinelearningmodelistrainedonaserverandsubsequentlydeployedacros...
FastDirectSearchinanOptimallyCompressedContinuousTargetSpaceforEfficientMulti-LabelActiveLearningWeishiShi1QiYu1Abstractretrievalandorganization.UsersfromQ&Awebsites,suchasstackoverflowandQuora,are...
DirectUncertaintyPredictionforMedicalSecondOpinionsMaithraRaghu12KatyBlumer2RorySayres2ZiadObermeyer3RobertKleinberg1SendhilMullainathan4JonKleinberg1Abstractlabellersnowbeinghighlytrainedmedicalex...
HighPerformanceZero-MemoryOverheadDirectConvolutionsJiyuanZhang1FranzFranchetti1TzeMengLow1AbstractPerformancenormalizedtoOpenBLASGEMMonAMDPileDriverThecomputationofconvolutionlayersindeep4.0GHz,4/...