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

Creation of the acceptable intelligent system based on the system intelligence circulation

Research Project

Project/Area Number 19H02159
Research Category

Grant-in-Aid for Scientific Research (B)

Allocation TypeSingle-year Grants
Section一般
Review Section Basic Section 21040:Control and system engineering-related
Research InstitutionYokohama National University

Principal Investigator

HAMAGAMI TOMOKI  横浜国立大学, 大学院工学研究院, 教授 (30334204)

Project Period (FY) 2019-04-01 – 2022-03-31
Project Status Completed (Fiscal Year 2021)
Budget Amount *help
¥17,160,000 (Direct Cost: ¥13,200,000、Indirect Cost: ¥3,960,000)
Fiscal Year 2021: ¥5,720,000 (Direct Cost: ¥4,400,000、Indirect Cost: ¥1,320,000)
Fiscal Year 2020: ¥5,720,000 (Direct Cost: ¥4,400,000、Indirect Cost: ¥1,320,000)
Fiscal Year 2019: ¥5,720,000 (Direct Cost: ¥4,400,000、Indirect Cost: ¥1,320,000)
Keywords知能システム / 機械学習 / 人工知能 / マルチドメイン学習 / 深層強化学習 / リザバーコンピューティング / サイバーフィジカルシステム / 説明可能なAI / 深層学習 / 転移学習 / 説明可能性 / 受容可能性
Outline of Research at the Start

近年、社会インフラ・製造・流通・サービスなど、あらゆる社会の仕組みの中で人工知能(AI)の活用による知能化・効率化が求められている。しかし、現在のAIはブラックボックス化されており、システムの理解と検証に必要な知識抽出が困難である。この問題に対し本研究では、説明可能なAI(eXplanable AI :XAI)の考え方を拡張した,システム知の循環に基づく知能システムの実現法を明らかにする。知能システム自らが,システム知を分析・説明・制御する機能とこれを循環する構造を有することで、XAIを超えた受容可能な知能システム基盤が創生されることを示す。

Outline of Final Research Achievements

These studies have achieved four main results toward the circulation mechanism of system knowledge for acceptable intelligent systems and system creation. First, we realized “adaptive threshold boosting” as a method for distilling knowledge extraction from experts and separating common and specific knowledge. Second, we clarified a method for adaptive partitioning and transferring multi-domain spaces by multi-agent approach for the advancement of machine learning valuation models in multi-domain spaces. Third, machine learning provides visual feedback and improved explanatory power to the shape design and refinement process. Furthermore, we realized the control problem by hierarchical reinforcement learning, and reused and circularized the intellectual structure by extracting its basic components.

Academic Significance and Societal Importance of the Research Achievements

サイバーフィジカルシステム(Cyber Physical Systems: CPS)は,今後の社会システムの姿を大きく変えるパラダイムである。CPSと人工知能の融合のためには,知能システム自らが,システム知を分析・説明・制御する仕組み持ち,この仕組みの循環を通して,XAIを超えた受容可能な知能システム(Acceptable Intelligent Systems: AIS)へと発展することが必要である。本研究はそのために必要なシステム知の抽出と説明法,設計と実装法,応用方法について,要素技術から具体的アプリケーションまで実現したことで,本分野の学術的および社会的に大きな意義を有する。

Report

(4 results)
  • 2021 Annual Research Report   Final Research Report ( PDF )
  • 2020 Annual Research Report
  • 2019 Annual Research Report
  • Research Products

    (24 results)

All 2022 2021 2020 2019

All Journal Article (10 results) (of which Peer Reviewed: 3 results) Presentation (14 results) (of which Int'l Joint Research: 3 results)

  • [Journal Article] Opticalflow Estimation for Deformable Object Utilizing Superpixels and Motion Estimation Using Phase Correlation and CNN2022

    • Author(s)
      Oishi Nobuya、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electronics, Information and Systems

      Volume: 142 Issue: 1 Pages: 100-107

    • DOI

      10.1541/ieejeiss.142.100

    • NAID

      130008139004

    • ISSN
      0385-4221, 1348-8155
    • Year and Date
      2022-01-01
    • Related Report
      2021 Annual Research Report
  • [Journal Article] Adversarial Inverse Reinforcement Learning to Estimate Policies from Multiple Experts2021

    • Author(s)
      Yamashita Kodai、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electronics, Information and Systems

      Volume: 141 Issue: 12 Pages: 1405-1410

    • DOI

      10.1541/ieejeiss.141.1405

    • NAID

      130008123513

    • ISSN
      0385-4221, 1348-8155
    • Year and Date
      2021-12-01
    • Related Report
      2021 Annual Research Report
  • [Journal Article] Learning Transferable Policy in Reinforcement Learning for Vehicle Velocity Tracking Control2021

    • Author(s)
      Natsu Yukiya、Hamagami Tomoki、Kanke Masayasu、Yoshida Kento、Niwakawa Makoto
    • Journal Title

      IEEJ Transactions on Electronics, Information and Systems

      Volume: 141 Issue: 12 Pages: 1492-1499

    • DOI

      10.1541/ieejeiss.141.1492

    • NAID

      130008123455

    • ISSN
      0385-4221, 1348-8155
    • Year and Date
      2021-12-01
    • Related Report
      2021 Annual Research Report
  • [Journal Article] Classification Model Considering Misclassification Based on Distributional Distance Between Labels2021

    • Author(s)
      Tajima Yoshiyuki、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electronics, Information and Systems

      Volume: 141 Issue: 9 Pages: 1048-1054

    • DOI

      10.1541/ieejeiss.141.1048

    • NAID

      130008082350

    • ISSN
      0385-4221, 1348-8155
    • Year and Date
      2021-09-01
    • Related Report
      2021 Annual Research Report
  • [Journal Article] Bayesian Network Oriented Transfer Learning Method for Credit Scoring Model2021

    • Author(s)
      Iwai Koichi、Akiyoshi Masanori、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electrical and Electronic Engineering

      Volume: 16 Issue: 9 Pages: 1195-1202

    • DOI

      10.1002/tee.23417

    • Related Report
      2021 Annual Research Report
  • [Journal Article] Reservoir Computing for Scalable Hardware with <scp>Block‐Based</scp> Neural Network2021

    • Author(s)
      Lee Kundo、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electrical and Electronic Engineering

      Volume: 16 Issue: 12 Pages: 1594-1602

    • DOI

      10.1002/tee.23473

    • Related Report
      2021 Annual Research Report
  • [Journal Article] ケアプラン作成支援システムのための非負値行列因子分解に基づく特徴語補完2021

    • Author(s)
      兵頭 幸起, 濱上 知樹
    • Journal Title

      情報処理学会論文誌

      Volume: 141 Pages: 369-377

    • NAID

      170000184256

    • Related Report
      2020 Annual Research Report
  • [Journal Article] Wave-making Resistance Estimation Through Deep Learning Considering the Distribution of Ship Figure2020

    • Author(s)
      Li Xin、Arai Hiroshi、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electronics, Information and Systems

      Volume: 140 Issue: 3 Pages: 391-397

    • DOI

      10.1541/ieejeiss.140.391

    • NAID

      130007804400

    • ISSN
      0385-4221, 1348-8155
    • Year and Date
      2020-03-01
    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Block-Based Neural Network Optimization with Manageable Problem Space2020

    • Author(s)
      Lee Kundo、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electronics, Information and Systems

      Volume: 140 Issue: 1 Pages: 68-74

    • DOI

      10.1541/ieejeiss.140.68

    • NAID

      130007779188

    • ISSN
      0385-4221, 1348-8155
    • Year and Date
      2020-01-01
    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Sperm Detection with Robustness to Overlap of Distributions by using Adaptive Thresholded Boosting2019

    • Author(s)
      Sasaki Hayato、Yamamoto Mizuki、Takeshima Teppei、Yumura Yasushi、Hamagami Tomoki
    • Journal Title

      IEEJ Transactions on Electronics, Information and Systems

      Volume: 139 Issue: 12 Pages: 1461-1467

    • DOI

      10.1541/ieejeiss.139.1461

    • NAID

      130007754486

    • ISSN
      0385-4221, 1348-8155
    • Year and Date
      2019-12-01
    • Related Report
      2019 Annual Research Report
    • Peer Reviewed
  • [Presentation] ファジィ測度を用いたアクティブラーニングによる効率的な識別学習2021

    • Author(s)
      木村慶豪, 濱上知樹
    • Organizer
      コンピューテーショナル・インテリジェンス研究会
    • Related Report
      2021 Annual Research Report
  • [Presentation] インスタンスの相互作用を用いたXGBoostによる不動産価格推定2021

    • Author(s)
      岩井康一, 濱上知樹
    • Organizer
      電気学会システム研究会
    • Related Report
      2021 Annual Research Report
  • [Presentation] Adversarial Inverse Reinforcement Learning to Estimate Policies from Multiple Experts2021

    • Author(s)
      Yamashita, Koudai, Hamagami,Tomoki
    • Organizer
      60th Annual Conference of the Society of Instrument and Control Engineers of Japan (SICE) 2021
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 不完全知覚環境におけるリザバーコンピューティングを用いた強化学習2021

    • Author(s)
      山下廣大, 濱上知樹
    • Organizer
      インテリジェントシステムシンポジウム(FAN2021)
    • Related Report
      2021 Annual Research Report
  • [Presentation] Ordinal Regression Based on the Distributional Distance Between Labels2021

    • Author(s)
      Tajima,Yasuyuki, Hamagami,Tomoki
    • Organizer
      IEEE Systems Man and Cybernetics Conference 2021 (SMC2021)
    • Related Report
      2021 Annual Research Report
    • Int'l Joint Research
  • [Presentation] 階層型強化学習と方策蒸留を用いた車両速度追従制御2021

    • Author(s)
      夏有輝也,濱上知樹,菅家正康,吉田健人,庭川誠
    • Organizer
      令和3年電気学会全国大会
    • Related Report
      2020 Annual Research Report
  • [Presentation] BERTを用いた医療文書からの固有表現抽出2021

    • Author(s)
      五井野琢也,濱上 知樹
    • Organizer
      第48回知能システムシンポジウム
    • Related Report
      2020 Annual Research Report
  • [Presentation] ファジィ測度を使用したアクティブラーニングによる物体識別2021

    • Author(s)
      木村慶豪, 濱上 知樹
    • Organizer
      第48回知能システムシンポジウム
    • Related Report
      2020 Annual Research Report
  • [Presentation] Structured Feature Derivation for Transfer Learning on Credit Scoring2020

    • Author(s)
      Koichi Iwai, Masanori Akiyoshi, Tomoki Hamagami
    • Organizer
      IEEE SMC2020
    • Related Report
      2020 Annual Research Report
    • Int'l Joint Research
  • [Presentation] エキスパート群から方策推定を行う敵対的逆強化学習2020

    • Author(s)
      山下廣大,濱上知樹
    • Organizer
      電気学会システム研究会
    • Related Report
      2020 Annual Research Report
  • [Presentation] 帯域別 pix2pix を用いた眼底異常検出2019

    • Author(s)
      藤井仁花, 濱上知樹
    • Organizer
      第29回インテリジェント・システム・シンポジウム
    • Related Report
      2019 Annual Research Report
  • [Presentation] 意味情報に基づく小規模ベイジアンネットワークを用いたケアマネジメントの評価値予測2019

    • Author(s)
      板橋雅弘,寺尾勇一,林慧子,佐野貴洋,竹林奈々子,濱上知樹
    • Organizer
      第29回インテリジェント・システム・シンポジウム
    • Related Report
      2019 Annual Research Report
  • [Presentation] CNN を用いた類似した三次元船型形状からの造波抵抗推定2019

    • Author(s)
      登内雅人, 新井洋, 濱上知樹
    • Organizer
      第18回情報科学技術フォーラム
    • Related Report
      2019 Annual Research Report
  • [Presentation] ランキング学習を用いた顕微授精時の精子選別尺度獲得の検討2019

    • Author(s)
      佐々木勇人,岸大輝,山本みずき,竹島徹平,湯村寧,濱上知樹
    • Organizer
      第18回情報科学技術フォーラム
    • Related Report
      2019 Annual Research Report

URL: 

Published: 2019-04-18   Modified: 2023-01-30  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi