Generalization in Deep Learning.pdf
Withadirectanalysisofneuralnetworks,thispaperpresentsamathematicallytight
generalizationtheorytopartiallyaddressanopenproblemregardingthegeneralization
ofdeeplearning.Unlikepreviousbound-basedtheory,ourmaintheoryisquantitatively
astightaspossibleforeverydata
下载地址
用户评论
