• Search Research Projects
  • Search Researchers
  • How to Use
  1. Back to previous page

Toward investigating the intrinsic mechanism of accelerated (sub)gradient methods for convex optimization problems

Research Project

Project/Area Number 18K11178
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 60020:Mathematical informatics-related
Research InstitutionTokyo Institute of Technology

Principal Investigator

Fukuda Mituhiro  東京工業大学, 情報理工学院, 准教授 (80334548)

Project Period (FY) 2018-04-01 – 2021-03-31
Project Status Completed (Fiscal Year 2020)
Budget Amount *help
¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2020: ¥390,000 (Direct Cost: ¥300,000、Indirect Cost: ¥90,000)
Fiscal Year 2019: ¥520,000 (Direct Cost: ¥400,000、Indirect Cost: ¥120,000)
Fiscal Year 2018: ¥650,000 (Direct Cost: ¥500,000、Indirect Cost: ¥150,000)
Keywords加速(劣)勾配法 / 凸最適化問題 / 一次法 / 停止条件 / 非凸最適化問題 / 近接法 / Bregman距離 / DC最適化問題 / 近接勾配法 / 凸関数 / 常微分方程式 / 加速勾配法
Outline of Final Research Achievements

The main results can be summarized in two.
First, we considered the minimization of convex functions including non-differentiable functions by the accelerated (sub)gradient methods. We proposed new methods that guarantee the convergence of the generated sequences when a more practical gradient mapping norm is used for the stopping criterion and also when the convex function satisfies a certain analytic inequality. As a result, we proved that the proposed method is a nearly optimal method.
The other result is the proposal of a new method that combines the Bregman distance and the proximity method for a minimization of non-convex functions. In particular, in the analysis of the DC function, which can be expressed as the difference of two convex functions, it was shown that the sequences generated by the proposed method converge to a stationary point under certain assumptions.

Academic Significance and Societal Importance of the Research Achievements

混沌としている凸最適化問題に対する加速(劣)勾配法の中でも現段階において勾配写像のノルムによる停止条件を考慮した場合の整理がある程度出来たと思っている。特に、この関数がパラメータに依存するHoelderian Error Boundを満たした時にそれらのパラメータを推定しながら更新ができる新たな手法の収束に関する解析が行えた。
また2つの凸関数の差として表せる非凸関数の最小化をBregman距離を用いた近接法を利用することにより既存の手法より緩い条件で収束を保証するものが提案できた。特に、正規分布に従う疑似乱数によって生成された位相回復に対して既存手法より格段に効率が良いことが確認できた。

Report

(4 results)
  • 2020 Annual Research Report   Final Research Report ( PDF )
  • 2019 Research-status Report
  • 2018 Research-status Report
  • Research Products

    (6 results)

All 2021 2019 2018

All Journal Article (1 results) (of which Peer Reviewed: 1 results) Presentation (5 results) (of which Int'l Joint Research: 2 results)

  • [Journal Article] Nearly Optimal First-Order Methods for Convex Optimization under Gradient Norm Measure: An Adaptive Regularization Approach2021

    • Author(s)
      Masaru Ito、Mituhiro Fukuda
    • Journal Title

      Journal of Optimization Theory and Applications

      Volume: 188 Issue: 3 Pages: 770-804

    • DOI

      10.1007/s10957-020-01806-7

    • NAID

      40022730610

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed
  • [Presentation] DC最適化問題に対するBregman距離を用いた近接アルゴリズム2021

    • Author(s)
      髙橋翔大、福田光浩、田中未来
    • Organizer
      日本オペレーションズ・リサーチ学会2021年春季研究発表会
    • Related Report
      2020 Annual Research Report
  • [Presentation] DC最適化問題に対するBregman距離を用いた近接アルゴリズムと複素最適化問題への拡張2021

    • Author(s)
      髙橋翔大、福田光浩、田中未来
    • Organizer
      研究集会「最適化:モデリングとアルゴリズム」
    • Related Report
      2020 Annual Research Report
  • [Presentation] Nearly optimal first-order method under Holderian error bound: An adaptive proximal point approach2019

    • Author(s)
      Masaru Ito and Mituhiro Fukuda
    • Organizer
      The Sixth International Conference on Continuous Optimization (ICCOPT2019)
    • Related Report
      2019 Research-status Report
    • Int'l Joint Research
  • [Presentation] 勾配のノルムを停止条件とする準最適な一次法2019

    • Author(s)
      伊藤勝, 福田光浩
    • Organizer
      日本オペレーションズ・リサーチ学会 2019 年春季研究発表会
    • Related Report
      2018 Research-status Report
  • [Presentation] An adaptive first order method for weakly smooth and uniformly convex problems2018

    • Author(s)
      Masaru Ito and Mituhiro Fukuda
    • Organizer
      The 23rd International Symposium on Mathematical Programming (ISMP 2018)
    • Related Report
      2018 Research-status Report
    • Int'l Joint Research

URL: 

Published: 2018-04-23   Modified: 2022-01-27  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi