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

Study on the rule representations and transparency of feature extracted images obtained by deep learning

Research Project

Project/Area Number 18K11481
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 61040:Soft computing-related
Research InstitutionMeiji University

Principal Investigator

Hayashi Yoichi  明治大学, 理工学部, 専任教授 (20189666)

Project Period (FY) 2018-04-01 – 2021-03-31
Project Status Completed (Fiscal Year 2020)
Budget Amount *help
¥4,420,000 (Direct Cost: ¥3,400,000、Indirect Cost: ¥1,020,000)
Fiscal Year 2020: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2019: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Fiscal Year 2018: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Keywordsディープラーニング / ルール抽出 / 医用画像 / 深層学習 / ブラックボックス / 解釈性 / 説明能力 / 名義変数 / 分かりやすさ / 透明性 / 解釈可能性 / 構造化データ / 放射線画像 / ルール表現 / 病理画像 / 予測モデル / アンサンブル学習 / 高精度分類器 / 説明可能AI / 透明化 / 画像特徴抽出
Outline of Final Research Achievements

Generally, it is often difficult to explain and/or the results obtained by deep learning. This situation is called "black box". Although, deep learning can be applicable to various applications, allowance for the black box is different. In this study, we focus medical images with strict accountability and proposed various rule representation such as rule extraction to enhance the accountability for the results obtained by deep learning for medical images such as DBN and CNN.

Academic Significance and Societal Importance of the Research Achievements

本研究の学術的な意義はディープラーニングを画像に対して適用する際、データセットが非構造化データセット(数値など)、非構造化データセット(コンピュータビジョン、産業応用画像)および非構造化データセット(医用画像)において学習方式および学習結果の説明能力と解釈性は大きく異なるので、それぞれに適したディープラーニングの方式を具体的に示した点にある。

この研究はクレジットスコアリング、peer-to-peerレンディング、材料画像、病理画像などの広範な応用範囲がありディープラーニングを核とした新しい人工知能システムの重要な起点になり日本における人工知能の研究および実応用を加速させる社会的な意義をもつ。

Report

(4 results)
  • 2020 Annual Research Report   Final Research Report ( PDF )
  • 2019 Research-status Report
  • 2018 Research-status Report
  • Research Products

    (24 results)

All 2020 2019 2018 Other

All Int'l Joint Research (6 results) Journal Article (16 results) (of which Int'l Joint Research: 3 results,  Peer Reviewed: 16 results,  Open Access: 13 results) Remarks (2 results)

  • [Int'l Joint Research] Czestochowa University of Technology(ポーランド)

    • Related Report
      2020 Annual Research Report
  • [Int'l Joint Research] Clark University(米国)

    • Related Report
      2020 Annual Research Report
  • [Int'l Joint Research] University of Geneva(スイス)

    • Related Report
      2019 Research-status Report
  • [Int'l Joint Research] Czestochowa University of Technology(ポーランド)

    • Related Report
      2019 Research-status Report
  • [Int'l Joint Research] Medical University of Graz(オーストリア)

    • Related Report
      2019 Research-status Report
  • [Int'l Joint Research] University of Geneva/Department of Computer Science(スイス)

    • Related Report
      2018 Research-status Report
  • [Journal Article] Does deep learning work well for categorical datasets with mainly nominal attributes?2020

    • Author(s)
      Yoichi Hayashi
    • Journal Title

      Electronics

      Volume: 9 Issue: 11 Pages: 1966-1966

    • DOI

      10.3390/electronics9111966

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] One-dimensional convolutional neural networks with feature selection for highly concise rule extraction from credit scoring datasets with heterogeneous attributes2020

    • Author(s)
      Yoichi Hayashi, Naoki Takano
    • Journal Title

      Electronics

      Volume: 9 Issue: 8 Pages: 1318-1318

    • DOI

      10.3390/electronics9081318

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Black Box Nature of Deep Learning for: Digital Pathology Beyond Quantitative to Qualitative Algorithmic Performances2020

    • Author(s)
      Yoichi Hayashi
    • Journal Title

      Lecture Notes on Artificial Intelligence (LNAI)

      Volume: 12090 Pages: 95-101

    • DOI

      10.1007/978-3-030-50402-1_6

    • ISBN
      9783030504014, 9783030504021
    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] Browser Fingerprint Coding Methods Increasing The Effectiveness of User Identification in The Web Traffic2020

    • Author(s)
      Marcin Gabryel, Konrad Grzanek, Yoichi Hayashi
    • Journal Title

      Journal of Artificial Intelligence and Soft Computing Research

      Volume: 10 Issue: 4 Pages: 243-253

    • DOI

      10.2478/jaiscr-2020-0016

    • Related Report
      2020 Annual Research Report
    • Peer Reviewed / Open Access
  • [Journal Article] New unified insights on deep learning in radiological and pathological images: beyond quantitative performances to qualitative interpretation2020

    • Author(s)
      Yoichi Hayashi
    • Journal Title

      Informatics in Medicine Unlocked

      Volume: 19 Pages: 100329-100329

    • DOI

      10.1016/j.imu.2020.100329

    • Related Report
      2020 Annual Research Report 2019 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Rough Support Vector Machine for Classification with Interval and Incomplete Data2020

    • Author(s)
      1.Nowicki Robert K, Grzanek K, Yoichi Hayashi
    • Journal Title

      Journal of Artificial Intelligence and Soft Computing Research

      Volume: 10 Pages: 47-56

    • Related Report
      2019 Research-status Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Journal Article] The right direction to be needed to develop a white-box deep learning in radiology, pathology, and ophthalmology: A short review2019

    • Author(s)
      Yoichi Hayashi
    • Journal Title

      Frontiers in Robotics and AI

      Volume: 6

    • DOI

      10.3389/frobt.2019.00024

    • Related Report
      2019 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Toward to transparency of deep learning in radiological imaging: Beyond quantitative to qualitative AI2019

    • Author(s)
      Yoichi Hayashi
    • Journal Title

      Journal of Medical Artificial Intelligence

      Volume: 2 Pages: 19-19

    • DOI

      10.21037/jmai.2019.09.06

    • Related Report
      2019 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Detection of lower albuminuria levels and early development of diabetic kidney disease using an artificial intelligence-based rule extraction approach2019

    • Author(s)
      Yoichi Hayashi
    • Journal Title

      Diagnostics

      Volume: 9 Pages: 33-33

    • Related Report
      2019 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Artificial intelligence-based rule extraction approach to derive the upper limit of hemoglobin during anemia treatment in patients with predialysis chronic kidney disease2019

    • Author(s)
      Yoichi Hayashi, Kei Nakajima, Keisuke Nakajima
    • Journal Title

      Informatics in Medicine Unlocked

      Volume: 17 Pages: 100262-100262

    • Related Report
      2019 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Optimality and convergence for convex ensemble learning with sparsity and diversity based on fixed point optimization2018

    • Author(s)
      Yoichi Hayashi and Hideaki Iiduka
    • Journal Title

      Neurocomputing

      Volume: 273 Pages: 367-372

    • DOI

      10.1016/j.neucom.2017.07.046

    • Related Report
      2018 Research-status Report
    • Peer Reviewed
  • [Journal Article] A comparison study on rule extraction from neural network ensembles, Boosted shallow trees and SVMs2018

    • Author(s)
      Guido Bologna, Yoichi Hayashi
    • Journal Title

      Applied Computional Intelligence and Soft Computing

      Volume: 2018 Pages: 1-20

    • DOI

      10.1155/2018/4084850

    • Related Report
      2018 Research-status Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Journal Article] High accuracy-priority rule extraction for reconciling accuracy an interpretability in credit scoring2018

    • Author(s)
      Yoichi Hayashi, Tatsuhiro Oisi
    • Journal Title

      New Generation Computing

      Volume: 36 Issue: 4 Pages: 393-418

    • DOI

      10.1007/s00354-018-0043-5

    • Related Report
      2018 Research-status Report
    • Peer Reviewed
  • [Journal Article] Non-invasive prediction of non-alcoholic steatohepatitis in Japanese patients with morbid obesity by artificial intelligence using rule extraction technology.2018

    • Author(s)
      Uehara D, Hayashi Y, Seki Y, Kakizaki S, Horiguchi N, Tojima H, Yamazaki Y, Sato K, Yasuda K, Yamada M, Uraoka T, Kasama K.
    • Journal Title

      World J Hepatol.

      Volume: Dec 27;10(12) Issue: 12 Pages: 934-943

    • DOI

      10.4254/wjh.v10.i12.934

    • Related Report
      2018 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Use of deep belief network for high-level abstraction small data sets using artificial intelligence with rule extraction2018

    • Author(s)
      Yoichi Hayashi
    • Journal Title

      Neural Computation

      Volume: 30 Issue: 12 Pages: 3309-3326

    • DOI

      10.1162/neco_a_01139

    • Related Report
      2018 Research-status Report
    • Peer Reviewed
  • [Journal Article] A Rule Extraction Study from SVM on Sentiment Analysis, Big Data and Cognitive Computing2018

    • Author(s)
      6.Guido Bologna, Yoichi Hayashi
    • Journal Title

      Big Data and Cognitive Computing

      Volume: 2018 Issue: 1 Pages: 6-6

    • DOI

      10.3390/bdcc2010006

    • Related Report
      2018 Research-status Report
    • Peer Reviewed / Open Access / Int'l Joint Research
  • [Remarks] 明治大学理工学部情報科学科人工知能研究室

    • URL

      http://www.ci.cs.meiji.ac.jp/

    • Related Report
      2020 Annual Research Report
  • [Remarks] 明治大学理工学部人工知能研究室

    • URL

      http://www.ci.cs.meiji.ac.jp/en/index.html

    • Related Report
      2018 Research-status Report

URL: 

Published: 2018-04-23   Modified: 2022-01-27  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi