Art

J-GLOBAL ID：201902233441564876   Reference number：19A0497363

DELAYED WEIGHT UPDATE FOR FASTER CONVERGENCE IN DATA-PARALLEL DEEP LEARNING

データ並列深層学習における高速収束のための遅延重み更新【JST・京大機械翻訳】

Author (7)： Youkawa Tetsuya (Graduate School of Science, Technology and Innovation, Kobe University, Kobe, Japan) ,  Mori Haruki (Graduate School of Science, Technology and Innovation, Kobe University, Kobe, Japan) ,  Miyauchi Yuki (Graduate School of Science, Technology and Innovation, Kobe University, Kobe, Japan) ,  Yamada Kazuki (Graduate School of Science, Technology and Innovation, Kobe University, Kobe, Japan) ,  Izumi Shintaro (Graduate School of Science, Technology and Innovation, Kobe University, Kobe, Japan) ,  Yoshimoto Masahiko (Graduate School of Science, Technology and Innovation, Kobe University, Kobe, Japan) ,  Kawaguchi Hiroshi (Graduate School of Science, Technology and Innovation, Kobe University, Kobe, Japan)
Material： IEEE Conference Proceedings  (IEEE Conf Proc)
Volume： 2018  Issue： GlobalSIP  Page： 663-667  Publication year： 2018 
JST Material Number： W2441A  Document type： Proceedings
Article type： 原著論文  Country of issue： United States (USA)  Language： ENGLISH (EN)

Abstract/Point： This paper presents a proposal...

Register(free) or Log In to access     To see more with JDream III (charged).   {{ this.onShowAbsJLink("http://jdream3.com/lp/jglobal/index.html?docNo=19A0497363&from=J-GLOBAL&jstjournalNo=W2441A") }}

Thesaurus term： telecommunication ,  machining time ,  data communication ,  neural network (biomodel) ,  network ,  queuing time ,  Communication ,  speedup ,  *high speed ,  deep learning ,  Time ,  machining condition ,  condition
Semi thesaurus term： *データ並列 , 【Automatic Indexing@JST】

JST classification (1)： Graphic and image processing in general  (JE04010I)

Terms in the title (5)： データ並列 ,  深層学習 ,  収束 ,  遅延 ,  更新