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

Natural reinforcement learning integrating intrinsic motivation and sociality

Research Project

Project/Area Number 20H04259
Research Category

Grant-in-Aid for Scientific Research (B)

Allocation TypeSingle-year Grants
Section一般
Review Section Basic Section 61040:Soft computing-related
Research InstitutionTokyo Denki University

Principal Investigator

Takahashi Tatsuji  東京電機大学, 理工学部, 教授 (00514514)

Co-Investigator(Kenkyū-buntansha) 甲野 佑  東京電機大学, 理工学部, 研究員 (10870313)
玉造 晃弘  東京電機大学, 理工学部, 研究員 (10876361)
太田 宏之  防衛医科大学校(医学教育部医学科進学課程及び専門課程、動物実験施設、共同利用研究施設、病院並びに防衛, 薬理学, 講師 (20535190)
浦上 大輔  日本大学, 生産工学部, 准教授 (40458196)
大用 庫智  関西学院大学, 総合政策学部, 講師 (60755685)
Project Period (FY) 2020-04-01 – 2023-03-31
Project Status Completed (Fiscal Year 2022)
Budget Amount *help
¥17,810,000 (Direct Cost: ¥13,700,000、Indirect Cost: ¥4,110,000)
Fiscal Year 2022: ¥5,200,000 (Direct Cost: ¥4,000,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2021: ¥5,200,000 (Direct Cost: ¥4,000,000、Indirect Cost: ¥1,200,000)
Fiscal Year 2020: ¥7,410,000 (Direct Cost: ¥5,700,000、Indirect Cost: ¥1,710,000)
Keywords強化学習 / 満足化 / 限定合理性 / 動物実験 / 機械学習 / 社会学習 / 因果推論 / 自然知能 / 天然知能 / 模倣 / バンディット問題 / 深層強化学習 / 採餌行動 / 行動経済学 / 模倣学習
Outline of Research at the Start

囲碁やビデオゲームなどで人間を上回る性能を見せている「人工強化学習」に対して、人間や動物の社会性や動機付け、環境の探索方法を組み込んだ「自然強化学習」を提案する。人工強化学習では、無数の致命的な失敗(=死)なしには学習が行えず、大量の個体集合で解決を図る。他方人間や多くの動物は、他個体の学習状況を観察して自らの内発的・外発的動機付けを調整し、結果、無駄な死も避ける。社会性を組み込んだ高効率な「自然強化学習」の理論とモデルを構築し、それを動物や人間の実験で検証するとともに、工学的な応用も実現する。

Outline of Final Research Achievements

aIn this project, we have formalized the mechanisms and merits of the natural reinforcement learning that humans and animals do. The formalization was done reconsidering the concepts of reward, motivation, task formalization (in terms of theory of computation), and sociality. Theoretically, we succeeded in a unification of bounded rationality, decision-making, and foraging theories from the notion of subjective regret. Some industrial applications were done and a principle of social learning under uncertainty was formulated. We also found that mice adaptively control the (asymmetric) learning rates under uncertainty, according to the environments that they face. It leads to a generalization of our theory.

Academic Significance and Societal Importance of the Research Achievements

人間や動物がどのように不確実な環境において学習しているかについての知見を深めました。これは今後、教育、訓練、社会活動などをどのように行うべきかについて指針を与える可能性があります。また、ChatGPTなどが人間と対話できるようにするために肝要な強化学習技術について、学習の目標を定めれば、それに向かって非常に効率的に学習を行えるようになりました。これは、生成AI、ゲーム技術、ロボット制御などにおいて広範な応用を得る可能性があります。

Report

(4 results)
  • 2022 Annual Research Report   Final Research Report ( PDF )
  • 2021 Annual Research Report
  • 2020 Annual Research Report
  • Research Products

    (23 results)

All 2023 2022 2021 2020

All Journal Article (9 results) (of which Int'l Joint Research: 2 results,  Peer Reviewed: 9 results,  Open Access: 9 results) Presentation (13 results) Book (1 results)

  • [Journal Article] Causal intuition in the indefinite world: Meta-analysis and simulations2023

    • Author(s)
      Higuchi Kohki、Oyo Kuratomo、Takahashi Tatsuji
    • Journal Title

      Biosystems

      Volume: 225 Pages: 104842-104842

    • DOI

      10.1016/j.biosystems.2023.104842

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Softsatisficing: Risk-sensitive softmax action selection2022

    • Author(s)
      Kamiya Takumi、Takahashi Tatsuji
    • Journal Title

      Biosystems

      Volume: 213 Pages: 104633-104633

    • DOI

      10.1016/j.biosystems.2022.104633

    • Related Report
      2021 Annual Research Report 2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Conveying Intention by Motions With Awareness of Information Asymmetry2022

    • Author(s)
      Fukuchi Yosuke、Osawa Masahiko、Yamakawa Hiroshi、Takahashi Tatsuji、Imai Michita
    • Journal Title

      Frontiers in Robotics and AI

      Volume: 9 Pages: 783863-783863

    • DOI

      10.3389/frobt.2022.783863

    • Related Report
      2021 Annual Research Report 2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Toward computational implementation of metaphor comprehension process based on the theory of indeterminate natural transformation2021

    • Author(s)
      池田 駿介、布山 美慕、西郷 甲矢人、高橋 達二
    • Journal Title

      Cognitive Studies: Bulletin of the Japanese Cognitive Science Society

      Volume: 28 Issue: 1 Pages: 39-56

    • DOI

      10.11225/cs.2020.065

    • NAID

      130007998460

    • ISSN
      1341-7924, 1881-5995
    • Year and Date
      2021-03-01
    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] The asymmetric learning rates of murine exploratory behavior in sparse reward environments2021

    • Author(s)
      Ohta Hiroyuki、Satori Kuniaki、Takarada Yu、Arake Masashi、Ishizuka Toshiaki、Morimoto Yuji、Takahashi Tatsuji
    • Journal Title

      Neural Networks

      Volume: 143 Pages: 218-229

    • DOI

      10.1016/j.neunet.2021.05.030

    • Related Report
      2021 Annual Research Report 2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Self-incremental learning vector quantization with human cognitive biases2021

    • Author(s)
      Manome Nobuhito、Shinohara Shuji、Takahashi Tatsuji、Chen Yu、Chung Ung-il
    • Journal Title

      Scientific Reports

      Volume: 11 Issue: 1 Pages: 3910-3910

    • DOI

      10.1038/s41598-021-83182-4

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] A category theoretic approach to metaphor comprehension: Theory of indeterminate natural transformation2020

    • Author(s)
      Fuyama Miho、Saigo Hayato、Takahashi Tatsuji
    • Journal Title

      Biosystems

      Volume: 197 Pages: 104213-104213

    • DOI

      10.1016/j.biosystems.2020.104213

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] A new method of Bayesian causal inference in non-stationary environments2020

    • Author(s)
      Shinohara Shuji、Manome Nobuhito、Suzuki Kouta、Chung Ung-il、Takahashi Tatsuji、Okamoto Hiroshi、Gunji Yukio Pegio、Nakajima Yoshihiro、Mitsuyoshi Shunji
    • Journal Title

      PLOS ONE

      Volume: 15 Issue: 5 Pages: e0233559-e0233559

    • DOI

      10.1371/journal.pone.0233559

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Journal Article] Extended Bayesian inference incorporating symmetry bias2020

    • Author(s)
      Shinohara Shuji、Manome Nobuhito、Suzuki Kouta、Chung Ung-il、Takahashi Tatsuji、Gunji Pegio-Yukio、Nakajima Yoshihiro、Mitsuyoshi Shunji
    • Journal Title

      Biosystems

      Volume: 190 Pages: 104104-104104

    • DOI

      10.1016/j.biosystems.2020.104104

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Presentation] 発話者間の関係性が条件推論の抑制に及ぼす影響 -ポライトネス理論に基づく検討-2023

    • Author(s)
      小倉那央,高橋達二,中村紘子
    • Organizer
      IPSJ2023
    • Related Report
      2022 Annual Research Report
  • [Presentation] 未来考慮型の信頼度に基づく合目的強化学習2023

    • Author(s)
      有村柊一,南 朱音,甲野 佑,高橋達二
    • Organizer
      IPSJ2023
    • Related Report
      2022 Annual Research Report
  • [Presentation] 複数の満足化エージェント間のわずかな情報共有による分業と最適性2022

    • Author(s)
      和田 拓真, 高橋 達二
    • Organizer
      JSAI2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 理想基準値を用いた確率的満足化方策2022

    • Author(s)
      加藤 暦雄, 甲野 佑, 高橋 達二
    • Organizer
      JSAI2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 信頼度を局所的に近似する認知的満足化方策2022

    • Author(s)
      南 朱音, 甲野 佑, 高橋 達二
    • Organizer
      JSAI2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] スケール可能かつシンプルな深層強化学習検証タスクの開発2022

    • Author(s)
      池田 龍司, 南 朱音, 甲野 佑, 高橋 達二
    • Organizer
      JSAI2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 不定自然変換理論の動的特性2022

    • Author(s)
      横須賀 天臣, 布山 美慕, 西郷 甲矢人, 高橋 達二
    • Organizer
      JSAI2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 確率的満足化における最適な基準値の動的推定2022

    • Author(s)
      久米 淳, 鈴木 裕毅, 加藤 暦雄, 甲野 祐, 高橋 達二
    • Organizer
      JSAI2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 希求水準の達成度合いを考慮する自然強化学習2022

    • Author(s)
      越川 駿平, 有村 柊一, 若林 洋尭, 甲野 佑, 高橋 達二
    • Organizer
      JSAI2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 思考スタイルの個人差が条件推論に及ぼす影響の検討2022

    • Author(s)
      横須賀 天臣, 渡邊 元樹, 高橋 達二, 中村 紘子
    • Organizer
      JCSS2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 人間の因果的直感に基づく因果探索アルゴリズム2022

    • Author(s)
      樋口 滉規, 高橋 達二
    • Organizer
      JCSS2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 追試とメタ分析による因果帰納推論モデルの適合性評価2022

    • Author(s)
      林 涼太, 市野 弘人, 樋口 滉規, 高橋 達二
    • Organizer
      JCSS2022
    • Related Report
      2022 Annual Research Report
  • [Presentation] 不定自然変換理論とWord2Vecを用いた比喩生成2022

    • Author(s)
      阿久津 規介, 池田 駿介, 布山 美慕, 西郷 甲矢人, 高橋 達二
    • Organizer
      JCSS2022
    • Related Report
      2022 Annual Research Report
  • [Book] Logic and Uncertainty in the Human Mind2020

    • Author(s)
      Takahashi Tatsuji、Oyo Kuratomo、Tamatsukuri Akihiro、Higuchi Kohki
    • Total Pages
      20
    • Publisher
      Routledge
    • Related Report
      2020 Annual Research Report

URL: 

Published: 2020-04-28   Modified: 2024-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi