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 1 1
    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