Geometrical formulation of the nonnegative matrix factorization

Shotaro Akaho, Hideitsu Hino, Neneka Nara, Noboru Murata

研究成果: Conference contribution

1 被引用数 (Scopus)

抄録

Nonnegative matrix factorization (NMF) has many applications as a tool for dimension reduction. In this paper, we reformulate the NMF from an information geometrical viewpoint. We show that a conventional optimization criterion is not geometrically natural, thus we propose to use more natural criterion. By this formulation, we can apply a geometrical algorithm based on the Pythagorean theorem. We also show the algorithm can improve the existing algorithm through numerical experiments.

本文言語English
ホスト出版物のタイトルNeural Information Processing - 25th International Conference, ICONIP 2018, Proceedings
編集者Long Cheng, Seiichi Ozawa, Andrew Chi Sing Leung
出版社Springer Verlag
ページ525-534
ページ数10
ISBN(印刷版)9783030041816
DOI
出版ステータスPublished - 2018
イベント25th International Conference on Neural Information Processing, ICONIP 2018 - Siem Reap, Cambodia
継続期間: 2018 12 132018 12 16

出版物シリーズ

名前Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
11303 LNCS
ISSN(印刷版)0302-9743
ISSN(電子版)1611-3349

Other

Other25th International Conference on Neural Information Processing, ICONIP 2018
CountryCambodia
CitySiem Reap
Period18/12/1318/12/16

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

フィンガープリント 「Geometrical formulation of the nonnegative matrix factorization」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