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

2011 Fiscal Year Final Research Report

A study on distance metric learning of data distribution space based on information theoretic criterion

Research Project

  • PDF
Project/Area Number 22800067
Research Category

Grant-in-Aid for Research Activity Start-up

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

Principal Investigator

HINO Hideitsu  早稲田大学, 理工学術院, 助教 (10580079)

Project Period (FY) 2010 – 2011
Keywords統計的学習理論
Research Abstract

We developed a method to optimize the distribution of features extracted from a given set of data. The features obtained by the method will follow a distribution, which is suitable for classification task. Though the method is applicable to any classification problem, we applied it to speaker recognition task, and achieved state-of-the-art performance. We also developed information estimators, which quantify the information contents in a datum. We applied the estimators to evaluate the credibility of prediction of solar power generated from solar panel.

  • Research Products

    (12 results)

All 2012 2011 2010

All Journal Article (6 results) (of which Peer Reviewed: 6 results) Presentation (6 results)

  • [Journal Article] Multiple Kernel Learning with Gaussianity Measures2012

    • Author(s)
      Hideitsu Hino, Nima Reyhani, Noboru Murata
    • Journal Title

      Neural Computation

      Volume: vol.24 Pages: 7

    • Peer Reviewed
  • [Journal Article] Extraction of Basic Patterns of Household Energy Consumption2011

    • Author(s)
      Haoyang Shen, Hideitsu Hino, Noboru Murata, Shinji Wakao
    • Journal Title

      International Conference on Machine Learning and Applications

      Pages: 275-280

    • Peer Reviewed
  • [Journal Article] Speaker verification robust to talking style variation using multiple kernel learning based on conditional entropy minimization2011

    • Author(s)
      Tetsuji Ogawa, Hideitsu Hino, Noboru Murata, Tetsunori Kobayashi
    • Journal Title

      INTERSPEECH

    • Peer Reviewed
  • [Journal Article] A Computationally Efficient Information Estimator for Weighted Data2011

    • Author(s)
      Hideitsu Hino, Noboru Murata
    • Journal Title

      Artificial Neural Networks and Machine Learning

      Volume: vol.6792 Pages: 301-308

    • Peer Reviewed
  • [Journal Article] Speaker recognition using multiple kernel learning based on conditional entropy minimization2011

    • Author(s)
      Tetsuji Ogawa, Hideitsu Hino, Noboru Murata, Tetsunori Kobayashi
    • Journal Title

      Acoustics, Speech and Signal Processing

      Pages: 2204-2207

    • Peer Reviewed
  • [Journal Article] Multiple Kernel Learning by Conditional Entropy Minimization2010

    • Author(s)
      Hideitsu Hino, Nima Reyhani, Noboru Murata
    • Journal Title

      International Conference on Machine Learning and Applications

      Pages: 223-228

    • Peer Reviewed
  • [Presentation] クラスタリングによる家庭消費電力パターンの抽出2011

    • Author(s)
      沈浩洋, 日野英逸, 村田昇
    • Organizer
      数理モデル化と問題解決研究会
    • Year and Date
      20111200
  • [Presentation] ガウス性に基づく多重カーネル学習2011

    • Author(s)
      日野英逸, 村田昇
    • Organizer
      第14回情報論的学習理論ワークショップ
    • Year and Date
      20111100
  • [Presentation] クラス内変動に頑健なカーネルマシンと話者照合への適用2011

    • Author(s)
      小川哲司, 日野英逸, 村田昇, 小林哲則
    • Organizer
      日本音響学会2011年秋季研究発表会
    • Year and Date
      20110900
  • [Presentation] 太陽光発電量予測とその信頼性評価2011

    • Author(s)
      村田昇, 沈浩洋, 日野英逸
    • Organizer
      日本鉄鋼協会第162回秋季講演大会
    • Year and Date
      20110900
  • [Presentation] 条件付きエントロピー最小化基準に基づくマルチカーネル学習を用いた発話スタイル変動に頑健な話者照合2011

    • Author(s)
      小川哲司, 日野英逸, 村田昇, 小林哲則
    • Organizer
      第87回音声言語情報処理研究会
    • Year and Date
      20110700
  • [Presentation] 分位点に基づく重み付きデータの情報量推定手法とその応用2010

    • Author(s)
      日野英逸, 村田昇
    • Organizer
      第13回情報論的学習理論ワークショップ
    • Year and Date
      20101100

URL: 

Published: 2013-07-31  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi