Accelerated stochastic multiplicative update with gradient averaging for nonnegative matrix factorizations

Hiroyuki Kasai*

*この研究の対応する著者

研究成果: Conference contribution

2 被引用数 (Scopus)

抄録

Nonnegative matrix factorization (NMF) is a powerful tool in data analysis by discovering latent features and part-based patterns from high-dimensional data, and is a special case in which factor matrices have low-rank nonnegative constraints. Applying NMF into huge-size matrices, we specifically address stochastic multiplicative update (MU) rule, which is the most popular, but which has slow convergence property. This present paper introduces a gradient averaging technique of stochastic gradient on the stochastic MU rule, and proposes an accelerated stochastic multiplicative update rule: SAGMU. Extensive computational experiments using both synthetic and real-world datasets demonstrate the effectiveness of SAGMU.

本文言語English
ホスト出版物のタイトル2018 26th European Signal Processing Conference, EUSIPCO 2018
出版社European Signal Processing Conference, EUSIPCO
ページ2593-2597
ページ数5
ISBN(電子版)9789082797015
DOI
出版ステータスPublished - 2018 11 29
外部発表はい
イベント26th European Signal Processing Conference, EUSIPCO 2018 - Rome, Italy
継続期間: 2018 9 32018 9 7

出版物シリーズ

名前European Signal Processing Conference
2018-September
ISSN(印刷版)2219-5491

Conference

Conference26th European Signal Processing Conference, EUSIPCO 2018
国/地域Italy
CityRome
Period18/9/318/9/7

ASJC Scopus subject areas

  • 信号処理
  • 電子工学および電気工学

フィンガープリント

「Accelerated stochastic multiplicative update with gradient averaging for nonnegative matrix factorizations」の研究トピックを掘り下げます。これらがまとまってユニークなフィンガープリントを構成します。

引用スタイル