A Fixed-Point Analysis of Regularized Dual Averaging under Static Scenarios

Masahiro Yukawa, Isao Yamada

研究成果: Conference contribution

抄録

In this paper, we analyze the properties of a fixed point of a certain mapping that is implicitly used in each of the regularized dual averaging (RDA) and projection-based RDA (PDA) algorithms. It turns out that, if the loss function has a nonexpansive (1-Lipschltz) gradient such as in the case of a half squared-distance function, RDA converges to a minimizer of the penalized loss function under a restrictive condition. Meanwhile, the fixed point for PDA gives a minimizer of the 'unpenalized' loss function. Some simulation studies are also presented to support the theoretical findings.

本文言語English
ホスト出版物のタイトル2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings
出版社Institute of Electrical and Electronics Engineers Inc.
ページ207-211
ページ数5
ISBN(電子版)9789881476852
DOI
出版ステータスPublished - 2019 3 4
イベント10th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Honolulu, United States
継続期間: 2018 11 122018 11 15

出版物シリーズ

名前2018 Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018 - Proceedings

Conference

Conference10th Asia-Pacific Signal and Information Processing Association Annual Summit and Conference, APSIPA ASC 2018
CountryUnited States
CityHonolulu
Period18/11/1218/11/15

ASJC Scopus subject areas

  • Information Systems

フィンガープリント 「A Fixed-Point Analysis of Regularized Dual Averaging under Static Scenarios」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