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

Algorithms and Applications of Nonnegative Matrix Factorization

Research Project

Project/Area Number 15K20986
Research Category

Grant-in-Aid for Young Scientists (B)

Allocation TypeMulti-year Fund
Research Field Social systems engineering/Safety system
Intelligent informatics
Research InstitutionShizuoka University (2018-2019)
Tokyo Institute of Technology (2015-2017)

Principal Investigator

Mizutani Tomohiko  静岡大学, 工学部, 講師 (00553984)

Project Period (FY) 2015-04-01 – 2020-03-31
Project Status Completed (Fiscal Year 2019)
Budget Amount *help
¥4,030,000 (Direct Cost: ¥3,100,000、Indirect Cost: ¥930,000)
Fiscal Year 2018: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2017: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2016: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2015: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Keywords非負行列分解 / 特異値分解 / 低ランク近似 / 逐次射影法 / 体積最小閉包楕円 / スペクトラル・クラスタリング / ハイパースペクトル画像 / 階層的交互最小二乗法 / 低ランク近似行列 / 前処理 / ノイズ頑強性 / 数理工学 / 機械学習
Outline of Final Research Achievements

This research project aimed to enhance algorithms for computing a nonnegative matrix factorization under separability assumption, and apply to extraction of features from big data. In particular, we examined the robustness of preconditioned successive projection algorithm, and improved the results shown by previous studies. Also, we developed an algorithm for computing singular value decompositions of matrices with low cost, and incorporated it into an ellipsoidal rounding algorithm and successive projection algorithm. We applied the resulting algorithms to endmember detection problems of hyperspectral images, and confirmed that they work well even if the data size is large.

Academic Significance and Societal Importance of the Research Achievements

分離可能性を仮定した下での非負行列分解は新聞記事を分類する問題やハイパースペクトラル画像から端成分を抽出する問題などに応用することができる.このような実問題への応用を想定したとき,計算手法はノイズに対して頑強で,大規模データに対しても動作することが望まれる.本研究課題では逐次射影法と楕円丸め法のアルゴリズムを改良することで,ノイズ頑強性の向上と計算コストの削減に成功した.そして,実際にハイパースペクトラル画像の端成分抽出問題に適用し,その有効性を確認した.以上の考察から開発した手法は大規模実データから有益な情報を抽出することに役立つことが期待できる.

Report

(6 results)
  • 2019 Annual Research Report   Final Research Report ( PDF )
  • 2018 Research-status Report
  • 2017 Research-status Report
  • 2016 Research-status Report
  • 2015 Research-status Report
  • Research Products

    (17 results)

All 2019 2018 2017 2016 2015

All Journal Article (8 results) (of which Open Access: 5 results,  Peer Reviewed: 2 results,  Acknowledgement Compliant: 1 results) Presentation (9 results) (of which Int'l Joint Research: 2 results)

  • [Journal Article] Improved Analysis of Spectral Algorithm for Clustering2019

    • Author(s)
      Tomohiko Mizutani
    • Journal Title

      arXiv:1912.02997

      Volume: -

    • NAID

      120006887330

    • Related Report
      2019 Annual Research Report
    • Open Access
  • [Journal Article] 凸計画に基づくスペクトラル・クラスタリング2019

    • Author(s)
      水谷友彦
    • Journal Title

      最適化:モデリングとアルゴリズム

      Volume: -

    • Related Report
      2018 Research-status Report
  • [Journal Article] Convex Programming Based Spectral Clustering2018

    • Author(s)
      Tomohiko Mizutani
    • Journal Title

      arXiv

      Volume: -

    • NAID

      120007053398

    • Related Report
      2017 Research-status Report
    • Open Access
  • [Journal Article] 楕円を用いたスペクトラル法の性能解析に向けて2018

    • Author(s)
      水谷友彦
    • Journal Title

      数理解析研究所講究禄

      Volume: 印刷中

    • Related Report
      2017 Research-status Report
  • [Journal Article] Efficient preconditioning for noisy separable nonnegative matrix factorization problems by successive projection based low-rank approximations2017

    • Author(s)
      Mizutani Tomohiko、Tanaka Mirai
    • Journal Title

      Machine Learning

      Volume: 107 Issue: 4 Pages: 643-673

    • DOI

      10.1007/s10994-017-5673-1

    • Related Report
      2017 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] 逐次射影法に基づく低ランク近似行列の計算2017

    • Author(s)
      水谷友彦,田中未来
    • Journal Title

      最適化:モデリングとアルゴリズム

      Volume: 29 Pages: 1-7

    • Related Report
      2016 Research-status Report
  • [Journal Article] Robustness analysis of preconditioned successive projection algorithm for general form of separable NMF problem2016

    • Author(s)
      Tomohiko Mizutani
    • Journal Title

      Linear Algebra and its Applications

      Volume: 497 Pages: 1-22

    • DOI

      10.1016/j.laa.2016.02.016

    • Related Report
      2015 Research-status Report
    • Peer Reviewed / Open Access / Acknowledgement Compliant
  • [Journal Article] 非負行列分解に対する前処理付き逐次射影法のノイズ頑強性2016

    • Author(s)
      水谷友彦
    • Journal Title

      数理解析研究所講究録

      Volume: 181 Pages: 45-50

    • Related Report
      2015 Research-status Report
    • Open Access
  • [Presentation] 凸計画に基づくスペクトラル・クラスタリング手法2018

    • Author(s)
      水谷友彦
    • Organizer
      科学研究費・基盤研究(A)「新時代の最適化モデルに基づく意思決定支援プラットフォームの研究と開発」による2018年度ワークショップ
    • Related Report
      2018 Research-status Report
  • [Presentation] 凸計画に基づくスペクトラル・クラスタリング2018

    • Author(s)
      水谷友彦
    • Organizer
      最適化:モデリングとアルゴリズム
    • Related Report
      2017 Research-status Report
  • [Presentation] 楕円を用いたスペクトラル・クラスタリングの性能評価2017

    • Author(s)
      水谷友彦
    • Organizer
      数理最適化の発展:モデル化とアルゴリズム
    • Related Report
      2017 Research-status Report
  • [Presentation] Efficient preconditioning for noisy separable nonnegative matrix factorization problems by successive projection based low-rank approximations2017

    • Author(s)
      Tomohiko Mizutani, Mirai Tanaka
    • Organizer
      The 9th Asian Conference on Machine Learning
    • Related Report
      2017 Research-status Report
    • Int'l Joint Research
  • [Presentation] 逐次射影法と部分空間反復法に基づく低ランク近似行列計算2016

    • Author(s)
      水谷友彦,田中未来
    • Organizer
      IBIS2016
    • Place of Presentation
      京都大学
    • Year and Date
      2016-11-16
    • Related Report
      2016 Research-status Report
  • [Presentation] 逐次射影法に基づく低ランク近似行列の計算2016

    • Author(s)
      水谷友彦,田中未来
    • Organizer
      最適化:モデリングとアルゴリズム
    • Place of Presentation
      政策研究大学院大学
    • Year and Date
      2016-03-22
    • Related Report
      2015 Research-status Report
  • [Presentation] 非負行列分解に対する前処理付き逐次射影法のノイズ頑強性2015

    • Author(s)
      水谷友彦
    • Organizer
      新時代を担う最適化:モデル化手法と数値計算
    • Place of Presentation
      京都大学
    • Year and Date
      2015-08-31
    • Related Report
      2015 Research-status Report
  • [Presentation] Spectral clustering by ellipsoid and its connection to separable nonnegative matrix factorization2015

    • Author(s)
      Tomohiko Mizutani
    • Organizer
      International Symposium on Optimization (ISMP)
    • Place of Presentation
      ピッツバーグ
    • Year and Date
      2015-07-15
    • Related Report
      2015 Research-status Report
    • Int'l Joint Research
  • [Presentation] 楕円を用いたスペクトラルクラスタリング2015

    • Author(s)
      水谷友彦
    • Organizer
      人工知能学会全国大会
    • Place of Presentation
      はこだて未来大学
    • Year and Date
      2015-06-02
    • Related Report
      2015 Research-status Report

URL: 

Published: 2015-04-16   Modified: 2022-11-04  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi