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

Generation of Orthogonal Sub-spaces for Efficient Learning in Layered Neural Networks with Asymmetric Structures

Research Project

Project/Area Number 20K11957
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Review Section Basic Section 61030:Intelligent informatics-related
Research InstitutionAdvanced Institute of Industrial Technology

Principal Investigator

Ishii Naohiro  東京都立産業技術大学院大学, 産業技術研究科, 研究員 (50004619)

Co-Investigator(Kenkyū-buntansha) 小田切 和也  椙山女学園大学, 文化情報学部, 教授 (30449491)
松尾 徳朗  東京都立産業技術大学院大学, 産業技術研究科, 教授 (80433142)
Project Period (FY) 2020-04-01 – 2023-03-31
Project Status Completed (Fiscal Year 2022)
Budget Amount *help
¥3,900,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥900,000)
Fiscal Year 2022: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2021: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000)
Fiscal Year 2020: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
Keywords非対称構造ネットワーク / Bio-inspired network / 非対称、非線形構造 / 直交基底の生成 / 高次直交基底の生成 / 多層構造の基底の生成 / 多層構造の高次基底の生成 / Bio-inspired ネットワーク / 直交部分空間の生成 / 非線形機能の空間生成 / 特徴空間でのトラッキング / 非対称構造の多層ネットワーク / 非対称構造のニューラルネット / Gabor filter / 選択的直交空間の生成 / 層状サブネットワークの構成 / 独立サブネットワークの設計 / 非対称構造でのGabor filter / 直交空間での特徴と学習 / 多層ニューラルネットの直交変換
Outline of Research at the Start

非対称構造を持つニューラルネットが適応性を持つ選択的直交空間の生成に寄与することを明らかにする。すでに、入力刺激の強度変化のある刺激に対して、ガボールフィルタを持つ非対称構造のネットワークの高い適応性を持つ直交化空間を実現できることを示した。われわれの研究では2次の非線形性を有する従来のEnergy modelの対称構造のネットワークよりも、非対称構造のネットワークが優れた方位選択性の能力のあることを示してきたが、さらに、直交空間の生成から局在性、スケール選択性、学習効率性・信頼性を向上した多層ネットの高い機能を持つ空間となることを明らかにする。

Outline of Final Research Achievements

In the orthogonal subspace of the visual system, nonlinear processing such as asymmetric structure and rectification is closely related. These two characteristics have been shown to generate orthogonal bases in the orthogonal subspace. We analyzed the tracking characteristics. This characteristic was shown to be superior to that of a symmetric model (called an Energy model) with conventional Gabor filters in an asymmetric neural network. Furthermore, it was proved on a vector space that the network with asymmetric structure has better classification ability than the target model. In addition, we took up the problem of generating higher-order orthogonal bases by multi-layer neural networks from lower-order bases.”

Academic Significance and Societal Importance of the Research Achievements

ニューラルネットワークの人工知能分野での適用が深層学習を中心として、大きく、進展している。しかし、深層のニューラルネットワークの処理のメカニズムの解明が十分でなく、ラックボックスでの処理として、残されている問題点も少なくない。そこで、本研究課題は明らかにしてきた生物の視覚神経系ネットワークをベースに、理解、説明可能な層状ネットワークの構成とその処理機構を明らかにして、深層学習のメカニズムの機能の基礎を明らかにすることである。

Report

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

    (12 results)

All 2023 2022 2021 2020

All Journal Article (12 results) (of which Peer Reviewed: 12 results,  Open Access: 1 results)

  • [Journal Article] Comparison of Fourier Bases and Asymmetric Networks in the Bio-inspired Networks2023

    • Author(s)
      Naohiro Ishii, Kazunori Iwata, Yuji Iwahori, Tokuro Matsuo
    • Journal Title

      Advances in Computational Intelligence - 18th International Work-Conference on Artificial Neural Networks, IWANN 2023

      Volume: LNCS(Springer) Pages: 1-12

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Classification Performance in the Bio-inspired Layered Networks2023

    • Author(s)
      Naohiro Ishii, Kazunori Iwata, Naoto Mukai, Kazuya Odagiri, Tokuro Matsuo
    • Journal Title

      International Congress on Information and Communication Technology (ICICT)

      Volume: LNNS(Springer) Pages: 1-12

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Generation of Orthogonality foe Feature Spaces in the Bio-inspired Neural Networks2022

    • Author(s)
      Naohiro Ishii, Toshinori Deguchi, Masashi Kawaguchi, Hiroshi Sasaki, Tokuro Matsuo
    • Journal Title

      Engineering Applications of Neural Networks-23rd EANN2022

      Volume: CCIS1600(Springer) Pages: 15-26

    • Related Report
      2022 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Tracking and Classification of Feature Space in the Bio-inspired Networks2022

    • Author(s)
      Naohiro Ishii, Toshinori Deguchi, Naoto Mukai, Kazuya Odagiri, Tokuro Matsuo
    • Journal Title

      Hybrid Artificial Intelligent Systems-17th HAIS

      Volume: LNCS13469(Springer) Pages: 27-38

    • DOI

      10.1007/978-3-031-15471-3_3

    • ISBN
      9783031154706, 9783031154713
    • Related Report
      2022 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Generation of Orthogonality for Feature Spaces in the Bio-inspired Neural Networks2022

    • Author(s)
      Naohiro Ishii, Toshinori Deguchi, Masashi Kawaguchi, Hiroshi Sasaki, Tokuro Matsuo
    • Journal Title

      Proceedings of the 23rd Engineering Applications of Neural Networks Conference - EANN 2022

      Volume: Editor Lazaros

    • Related Report
      2021 Research-status Report
    • Peer Reviewed
  • [Journal Article] Adaptive Orthogonal Characteristics of Bio-Inspired Neural Networks2021

    • Author(s)
      Ishii Naohiro、Deguchi Toshinori、Kawaguchi Masashi、Sasaki Hiroshi、Matsuo Tokuro
    • Journal Title

      Logic Journal of the IGPL

      Volume: JNL Issue: 4 Pages: 1-21

    • DOI

      10.1093/jigpal/jzab004

    • Related Report
      2022 Annual Research Report 2021 Research-status Report 2020 Research-status Report
    • Peer Reviewed / Open Access
  • [Journal Article] Generation of Chow Parameters and Reduced Variables Through Nearest Neighbor Relations in Threshold Networks2021

    • Author(s)
      Ishii Naohiro、Matsuo Tokuro
    • Journal Title

      International Journal of Neural Systems

      Volume: 31 Issue: 10 Pages: 2150045-2150045

    • DOI

      10.1142/s0129065721500453

    • Related Report
      2021 Research-status Report
    • Peer Reviewed
  • [Journal Article] Inheritances of Orthogonality in the Bio-inspired Layered Networks2021

    • Author(s)
      Ishii Naohiro、Deguchi Toshinori、Kawaguchi Masashi、Sasaki Hiroshi、Matsuo Tokuro
    • Journal Title

      Intelligent Data Engineering and Automated Learning -IDEAL2021

      Volume: LNCS 13113 Pages: 21-32

    • DOI

      10.1007/978-3-030-91608-4_3

    • ISBN
      9783030916077, 9783030916084
    • Related Report
      2021 Research-status Report
    • Peer Reviewed
  • [Journal Article] Features Spaces with Reduced Variables Based on Nearest Neighbor Relations and?Their Inheritances2021

    • Author(s)
      Ishii Naohiro、Iwata Kazunori、Mukai Naoto、Odagiri Kazuya、Matsuo Tokuro
    • Journal Title

      Advances in Computational Intellgence, IWANN2021

      Volume: LNCS 12861 Pages: 77-88

    • DOI

      10.1007/978-3-030-85030-2_7

    • ISBN
      9783030850296, 9783030850302
    • Related Report
      2021 Research-status Report
    • Peer Reviewed
  • [Journal Article] Reduction of Variables through Nearest Neighbor Relations in Threshold Networks2020

    • Author(s)
      Naohiro Ishii, Kazunori Iwata, Kazuya, Odagiri, Toyoshiro Nakashima, Tokuro Matsuo
    • Journal Title

      International Journal of Smart Computing and Artificial Intelligence

      Volume: 4(1) Pages: 36-54

    • Related Report
      2020 Research-status Report
    • Peer Reviewed
  • [Journal Article] Reduction of Variables and Generation of Functions Through Nearest Neighbor Relations in Threshold Networks2020

    • Author(s)
      Naohiro Ishii, Kazunori Iwata, Kazuya Odagiri, Toyoshiro Nakashima, Tokuro Matsuo
    • Journal Title

      Proc. of the 21st Engineering Application of Neural Networks, Springer

      Volume: 2 Pages: 569-578

    • DOI

      10.1007/978-3-030-48791-1_45

    • ISBN
      9783030487904, 9783030487911
    • Related Report
      2020 Research-status Report
    • Peer Reviewed
  • [Journal Article] Tourists Movement Analysis Based on Entropies of Markov Process2020

    • Author(s)
      Naohiro Ishii, Kazuya Odagiri, Hidekazu Iwamoto, Satoshi Takahashi, Kazunori Iwata, Tokuro Matsuo
    • Journal Title

      HAIS2020: Hybrid Artificial Intelligent Systems, LNCS, vol.12344, Springer

      Volume: 12344 Pages: 573-584

    • Related Report
      2020 Research-status Report
    • Peer Reviewed

URL: 

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

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi