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

Deriving Appropriate Utility in Agent Learning

Research Project

Project/Area Number 18700145
Research Category

Grant-in-Aid for Young Scientists (B)

Allocation TypeSingle-year Grants
Research Field Intelligent informatics
Research InstitutionOsaka University

Principal Investigator

MORIYAMA Koichi  Osaka University, 産業科学研究所, 助教 (10361776)

Project Period (FY) 2006 – 2008
Project Status Completed (Fiscal Year 2008)
Budget Amount *help
¥2,510,000 (Direct Cost: ¥2,300,000、Indirect Cost: ¥210,000)
Fiscal Year 2008: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2007: ¥800,000 (Direct Cost: ¥800,000)
Fiscal Year 2006: ¥800,000 (Direct Cost: ¥800,000)
Keywords人工知能 / 強化学習 / エージェント / マルチエージェントシステム / ゲーム理論 / 効用 / Q学習 / 囚人のジレンマゲーム
Research Abstract

本研究課題は,人工知能研究の一分野である強化学習において,従来は同一視されてきた個体(エージェント)外からの報酬と,エージェント自身の主観的効用を分けて考えることで,最も単純なマルチエージェント環境である2人2行動同時手番対称ゲームで適切に行動する強化学習エージェントの構築を目的とした.研究の結果,囚人のジレンマゲームにおいて協調行動を続けやすくする効用の条件を導き,さらに別の種類のゲームで報酬を追求することを妨げない学習手法を開発した.

Report

(4 results)
  • 2008 Annual Research Report   Final Research Report ( PDF )
  • 2007 Annual Research Report
  • 2006 Annual Research Report
  • Research Products

    (20 results)

All 2009 2008 2007 2006

All Journal Article (20 results) (of which Peer Reviewed: 15 results)

  • [Journal Article] Learning-Rate Adjust-ing Q-learning for Two-Person Two-Action Symmetric Games2009

    • Author(s)
      Koichi Moriyama
    • Journal Title

      Proceedings of the third KES Symposium on Agent and Multi-Agent Systems - Technologies and Applications, KES-AMSTA 2009 (Lecture Notes in Artificial Intelligence 5559) in press

      Pages: 223-232

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] Learning-Rate Adjust-ing Q-learning for Prisoner's Dilemma Games2008

    • Author(s)
      Koichi Moriyama
    • Journal Title

      Proceedings of the 2008 IEEE/ WIC/ACM International Conference on Intelligent Agent Technology IAT'08

      Pages: 322-325

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] 2人2行動ゲームのための学習率調整Q学習2008

    • Author(s)
      森山甲一
    • Journal Title

      合同エージェントワークショップ&シンポジウム2008 (JAWS2008) 論文集,

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] 囚人のジレンマゲームにおけるQ学習による協調の維持2008

    • Author(s)
      森山甲一.
    • Journal Title

      コンピュータソフトウェア Vol. 25, No. 4

      Pages: 145-153

    • NAID

      130004892116

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] Reinforcement Learn-ing on a Futures Market Simulator2008

    • Author(s)
      Koichi Moriyama, Mitsuhiro Matsumoto, Ken-ichi Fukui, Satoshi Kurihara, and Masayuki Numao
    • Journal Title

      Jour-nal of Universal Computer Science Vol. 14, No. 7

      Pages: 1136-1153

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] Reinforcement Learning on a Futures Market Simulator2008

    • Author(s)
      K. Moriyama, M. Matsumoto, K. Fukui, S. Kurihara, and M. Numao
    • Journal Title

      Journal of Universal Computer Science 14

      Pages: 1136-1153

    • Related Report
      2008 Annual Research Report
    • Peer Reviewed
  • [Journal Article] 囚人のジレンマゲームにおけるQ学習による協調の維持2008

    • Author(s)
      森山甲一
    • Journal Title

      コンピュータソフトウェア 25

      Pages: 145-153

    • NAID

      130004892116

    • Related Report
      2008 Annual Research Report
    • Peer Reviewed
  • [Journal Article] 2人2行動ゲームのための学習率調整Q学習2008

    • Author(s)
      森山甲一
    • Journal Title

      合同エージェントワークショップ&シンポジウム2008講演論文集

    • Related Report
      2008 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Learning-Rate Adjusting Q-learning for Prisoner's Dilemma Games2008

    • Author(s)
      Koichi Moriyama
    • Journal Title

      Proc. of the 2008 IEEE/WIC/ACM International Conference on Intelligent Agent Technology

      Pages: 322-325

    • Related Report
      2008 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Utility Based Q-learn-ing to Maintain Cooperation in Prisoner's Dilemma Games2007

    • Author(s)
      Koichi Moriyama
    • Journal Title

      Proceedings of the 2007 IEEE/WIC/ACM International Conference on Intelligent Agent Technology IAT'07

      Pages: 146-152

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] 囚人のジレンマゲームにおけるQ学習による協調の維持2007

    • Author(s)
      森山甲一
    • Journal Title

      合同エージェントワークショップ&シンポジウム2007 (JAWS2007)論文集

    • NAID

      130004892116

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] 囚人のジレンマゲームにおけるQ学習による協調の維持2007

    • Author(s)
      森山甲一
    • Journal Title

      第6回情報科学技術フォーラム (FIT2007) 講演論文集

      Pages: 419-422

    • NAID

      130004892116

    • Related Report
      2008 Final Research Report
  • [Journal Article] Reinforcement Learn-ing on a Futures Market Simulator2007

    • Author(s)
      Koichi Moriyama, Mitsuhiro Matsumoto, Ken-ichi Fukui, Satoshi Kurihara, and Masayuki Numao
    • Journal Title

      Proceedings of the first KES Symposium on Agent and Multi-Agent Systems - Technolo-gies and Applications, KES-AMSTA 2007 (Lecture Notes in Artificial Intelligence 4496)

      Pages: 42-52

    • Related Report
      2008 Final Research Report
    • Peer Reviewed
  • [Journal Article] Reinforcement Learning on a Futures Market Simulator2007

    • Author(s)
      K. Moriyama, M. Matsumoto, K. Fukui, S. Kurihara, and M. Numao
    • Journal Title

      Lecture Notes in Artificial Intelligence 4496

      Pages: 42-52

    • Related Report
      2007 Annual Research Report
    • Peer Reviewed
  • [Journal Article] 囚人のジレンマゲームにおけるQ学習による協調の維持2007

    • Author(s)
      森山 甲一
    • Journal Title

      第6回情報科学技術フォーラム講演論文集

      Pages: 419-422

    • NAID

      130004892116

    • Related Report
      2007 Annual Research Report
  • [Journal Article] 囚人のジレンマゲームにおけるQ学習による協調の維持2007

    • Author(s)
      森山 甲一
    • Journal Title

      合同エージェントワークショップ&シンポジウム2007講演論文集

    • NAID

      130004892116

    • Related Report
      2007 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Utility Based Q-learning to Maintain Cooperation Prisoner's Dilemma Games2007

    • Author(s)
      Koichi Moriyama
    • Journal Title

      Proc. 2007 IEEE/WIC/ACM International Conference on Intelligent Agent Technology

      Pages: 146-152

    • Related Report
      2007 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Reinforcement Learning on a Futures Market Simulator2007

    • Author(s)
      K.Moriyama, M.Matsumoto, K.Fukui, S.Kurihara, M.Numao
    • Journal Title

      Lecture Notes in Artificial Intelligence (in press)

    • Related Report
      2006 Annual Research Report
  • [Journal Article] U-MartにおけるQ学習エージェントの設計と評価2006

    • Author(s)
      松本光弘, 福井健一, 森山甲一, 栗原聡, 沼尾正行
    • Journal Title

      人工知能学会全国大会(第20回)論文集 1B2-2

    • NAID

      130005023261

    • Related Report
      2008 Final Research Report
  • [Journal Article] U-MartにおけるQ学習工ージェントの設計と評価2006

    • Author(s)
      松本光弘, 福井健一, 森山甲一, 栗原聡, 沼尾正行
    • Journal Title

      人工知能学会全国大会(第20回)論文集

    • Related Report
      2006 Annual Research Report

URL: 

Published: 2006-04-01   Modified: 2016-04-21  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi