ADifferentiablePointProcesswithitsApplicationtoSpikingNeuralNetworksHiroshiKajino1Abstractraisesthefollowingquestion:howmuchofthecurrentper-formancegapscanbeattributedtothisdifferenceonneuralThispa...
It’sNotWhatMachinesCanLearn,It’sWhatWeCannotTeachGalYehuda1MosheGabel2AssafSchuster1AbstractdatasetgenerationprocesscomplexityofresultingtaskCandeepneuralnetworkslearntosolveanytask,harderandinpa...
SubmodularStreaminginAllitsGlory:TightApproximation,MinimumMemoryandLowAdaptiveComplexityEhsanKazemi1MarkoMitrovic1MortezaZadimoghaddam2SilvioLattanzi2AminKarbasi1Abstractnon-negativesetfunctionf:2...
OnlineAdaptivePrincipalComponentAnalysisanditsextensionsJianjunYuan1AndrewLamperski1AbstractThestandardperformancemetricforonlinelearningmea-suresthedifferencebetweenthedecisionmaker’scumula-Wepro...
MolecularHypergraphGrammarwithitsApplicationtoMolecularOptimizationHiroshiKajino1Abstractproblemistoobtainm∈Msuchthat,Molecularoptimizationaimstodiscovernovelm=argmaxf(m),(1)moleculeswithdesirable...
ErrorCompensatedQuantizedSGDanditsApplicationstoLarge-scaleDistributedOptimizationJiaxiangWu1WeidongHuang1JunzhouHuang1TongZhang1AbstractTherehavebeenseveralworksattemptingtoimprovetheefficiencyofd...