Information geometric perspective of modal linear regression

Keishi Sando, Shotaro Akaho, Noboru Murata, Hideitsu Hino

研究成果: Conference contribution

1 被引用数 (Scopus)

抄録

Modal linear regression (MLR) is a standard method for modeling the conditional mode of a response variable using a linear combination of explanatory variables. It is effective when dealing with response variables with an asymmetric, multi-modal distribution. Because of the nonparametric nature of MLR, it is difficult to construct a statistical model manifold in the sense of information geometry. In this work, a model manifold is constructed using observations instead of explicit parametric models. We also propose a method for constructing a data manifold based on an empirical distribution. The em algorithm, which is a geometric formulation of the EM algorithm, of MLR is shown to be equivalent to the conventional EM algorithm of MLR.

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535-545
ページ数11
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 「Information geometric perspective of modal linear regression」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