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

Development of method to analyze the high-dimensional and small-sample data based on machine learning and its application to cryptanalysis

Research Project

Project/Area Number 21700308
Research Category

Grant-in-Aid for Young Scientists (B)

Allocation TypeSingle-year Grants
Research Field Statistical science
Research InstitutionKyushu University

Principal Investigator

KAWAKITA Masanori  九州大学, システム情報科学研究院・情報学部門 (90435496)

Project Period (FY) 2009 – 2011
Project Status Completed (Fiscal Year 2011)
Budget Amount *help
¥3,250,000 (Direct Cost: ¥2,500,000、Indirect Cost: ¥750,000)
Fiscal Year 2011: ¥780,000 (Direct Cost: ¥600,000、Indirect Cost: ¥180,000)
Fiscal Year 2010: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2009: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Keywords統計的学習理論 / n<<p問題 / 変数選択 / 重み付き尤度法 / 半教師付き学習 / 統計的パラドックス / 密度比推定 / n≪p問題
Research Abstract

It is already known that the estimation accuracy of supervised learning can be improved by using the unlabeled data even when the number of labeled data is quite small. This type of learning is called semi-supervised learning. The most conventional semi-supervised learning requires some additional assumptions to dominate the supervised learning even though we have additional information. Further, as for model selection, the conventional criteria(including AIC or CV) are applied to the labeled data. However, because such criteria require a large number of labeled data, they do not work well in this setting. Our main result is that we solved these problems.

Report

(4 results)
  • 2011 Annual Research Report   Final Research Report ( PDF )
  • 2010 Annual Research Report
  • 2009 Annual Research Report
  • Research Products

    (23 results)

All 2011 2010 2009

All Journal Article (2 results) (of which Peer Reviewed: 2 results) Presentation (21 results)

  • [Journal Article] Acceleration technique for boosting classification and its application to face detection2011

    • Author(s)
      M.Kawakita, R.Izumi, J.Takeuchi, Y.Hu, T.Takamori, H.Kameyama
    • Journal Title

      JMLR W&CP

      Volume: 20 Pages: 335-349

    • Related Report
      2011 Annual Research Report
    • Peer Reviewed
  • [Journal Article] A note on model selection for small sample regression2010

    • Author(s)
      Kawakita, M.Oie, Y.Takeuchi, J.
    • Journal Title

      Proceedings of International Symposium on Information Theory and its Applications 2010

      Pages: 112-117

    • Related Report
      2010 Annual Research Report
    • Peer Reviewed
  • [Presentation] A class of semi-supervised regression based on density ratio estimation2011

    • Author(s)
      Kawakita, M.
    • Organizer
      ENSEEIHT-Kyushu University Workshop on Data Mining and Media Processing
    • Place of Presentation
      ENSEEIHT
    • Related Report
      2011 Final Research Report
  • [Presentation] 安全な半教師付き回帰のクラスとそのモデル選択2011

    • Author(s)
      川喜田雅則
    • Organizer
      ミニワークショップ統計多様体の幾何学とその周辺(3)/幾何学と諸科学の連携調査
    • Place of Presentation
      北海道大学
    • Related Report
      2011 Final Research Report
  • [Presentation] ブースティングの判別高速化と顔検出への応用2011

    • Author(s)
      泉亮多, 川喜由雅則, 竹内純一, 胡軼, 高森哲弥, 亀山祐和
    • Organizer
      第6回IBISML研究会
    • Place of Presentation
      公立はこだて未来大学
    • Related Report
      2011 Annual Research Report
  • [Presentation] ベイズ符号長による確率的コンプレキシティの評価とモデル選択への応用2011

    • Author(s)
      武石啓成, 川喜田雅則, 竹内純一
    • Organizer
      第6回IBISML研究会
    • Place of Presentation
      公立はこだて未来大学
    • Related Report
      2011 Annual Research Report
  • [Presentation] サンプル間の相関を考慮した主成分分析とカーネル関数による共分散行列の推定2011

    • Author(s)
      林浩平, 川喜田雅則, 池田和司
    • Organizer
      IBIS 2011
    • Place of Presentation
      奈良女子大学
    • Related Report
      2011 Annual Research Report
  • [Presentation] A class of semi-supervised regression based on density ratioestimation2011

    • Author(s)
      Kawakita, M.
    • Organizer
      ENSEEIHT-Kyushu University Workshop on Data Mining and Media Processing
    • Place of Presentation
      ENSEEIHT, Toulouse
    • Related Report
      2011 Annual Research Report
  • [Presentation] ラベル無しデータを利用した回帰の改良2010

    • Author(s)
      川喜田雅則,竹内純一
    • Organizer
      第十二回人口知能学会データマイニングと統計数理研究会(SIG-DMSM)
    • Place of Presentation
      統計数理研究所
    • Related Report
      2011 Final Research Report
  • [Presentation] Semi-supervised learning in view of estimating functions2010

    • Author(s)
      Kawakita, M. and Takeuchi, J.
    • Organizer
      Information Geometry And Its Applications III
    • Related Report
      2011 Final Research Report
  • [Presentation] A note on model selection for small sample regression2010

    • Author(s)
      Kawakita, M., Oie, Y. and Takeuchi, J.
    • Organizer
      International Symposium on Information Theory and its Applications 2010
    • Place of Presentation
      Taichung, Taiwan
    • Related Report
      2011 Final Research Report
  • [Presentation] 大標本仮定を必要としない半教師付き回帰のモデル選択2010

    • Author(s)
      川喜田雅則,竹内純一
    • Organizer
      第13回情報論的学習理論ワークショップ
    • Place of Presentation
      東京大学駒場キャンパス
    • Related Report
      2011 Final Research Report
  • [Presentation] 情報幾何学と統計多様体上の一般化共形構造の周辺密度比推定を用いた半教師付き回帰法の改良2010

    • Author(s)
      川喜田雅則,竹内純一
    • Organizer
      情報幾何学と統計多様体上の一般化共形構造の周辺
    • Place of Presentation
      東北学院大学多賀城キャンパス
    • Related Report
      2011 Final Research Report
  • [Presentation] 半教師付き回帰のためのモデル選択2010

    • Author(s)
      川喜田雅則
    • Organizer
      情報幾何学と統計多様体上の一般化共形構造の周辺
    • Place of Presentation
      東北学院大学多賀城キャンパス
    • Related Report
      2011 Final Research Report
  • [Presentation] Semi-supervised learning in view of estimating functions2010

    • Author(s)
      Kawakita, M., Takeuchi, J.
    • Organizer
      Information Geometry And Its Applications III
    • Place of Presentation
      University of Leipzig
    • Related Report
      2010 Annual Research Report
  • [Presentation] 大標本仮定を必要としない半教師付き回帰のモデル選択2010

    • Author(s)
      川喜田雅則, 竹内純一
    • Organizer
      第13回情報論的学習理論ワークショップ(IBIS)
    • Place of Presentation
      東京大学駒場キャンパス
    • Related Report
      2010 Annual Research Report
  • [Presentation] 密度比推定を用いた半教師付き回帰法の改良2010

    • Author(s)
      川喜田雅則, 竹内純一
    • Organizer
      情報理論とその応用シンポジウム2010
    • Place of Presentation
      信州松代ロイヤルホテル
    • Related Report
      2010 Annual Research Report
  • [Presentation] 情報幾何によるブースティングの性能の考察2009

    • Author(s)
      川喜田雅則
    • Organizer
      データ科学特別セミナー
    • Place of Presentation
      大阪大学
    • Year and Date
      2009-07-15
    • Related Report
      2011 Final Research Report
  • [Presentation] ベイズ推定を用いない曲指数型分布族の推定量の改善2009

    • Author(s)
      川喜田雅則,竹内純一
    • Organizer
      第12回情報論的学習理論ワークショップ(IBIS 2009)
    • Place of Presentation
      九州大学(病院地区百年講堂)
    • Related Report
      2011 Final Research Report
  • [Presentation] A new method of improving predictive distribution without Bayes estimation2009

    • Author(s)
      Kawakita, M. and Takeuchi, J.
    • Organizer
      情報理論とその応用シンポジウム2009
    • Place of Presentation
      山口県湯田温泉ホテルかめ福
    • Related Report
      2011 Final Research Report
  • [Presentation] ベイズ推定を用いない曲指数型分布族の推定量の改善2009

    • Author(s)
      川喜田雅則, 竹内純一
    • Organizer
      第12回情報論的学習理論ワークショップ
    • Place of Presentation
      九州大学(百年講堂)
    • Related Report
      2009 Annual Research Report
  • [Presentation] A new method of improving predictive distribution without Bayes estimation2009

    • Author(s)
      Kawakita, M., Takeuchi, J.
    • Organizer
      情報理論とその応用学会2009
    • Place of Presentation
      山口県 湯田温泉
    • Related Report
      2009 Annual Research Report
  • [Presentation] ラベル無しデータを利用した回帰の改良2009

    • Author(s)
      川喜田雅則, 竹内純一
    • Organizer
      第12回データマイニングと統計数理研究会
    • Place of Presentation
      統計数理研究所
    • Related Report
      2009 Annual Research Report

URL: 

Published: 2009-04-01   Modified: 2016-04-21  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi