Matching Networks for One Shot Learning.pdf
Abstract:Learningfromafewexamplesremainsakeychallengeinmachinelearning.Despiterecentadvancesinimportantdomainssuchasvisionandlanguage,thestandardsuperviseddeeplearningparadigmdoesnotofferasatisfactorysolutionforlearningnewconceptsrapidlyfromlittledata.Inthiswork,weemployideasfrommet
下载地址
用户评论
