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

A Training Algorithm based on Gradient Method for Big Data Including High-nonlinearity

Research Project

Project/Area Number 26330287
Research Category

Grant-in-Aid for Scientific Research (C)

Allocation TypeMulti-year Fund
Section一般
Research Field Soft computing
Research InstitutionShonan Institute of Technology

Principal Investigator

Ninomiya Hiroshi  湘南工科大学, 工学部, 教授 (60308335)

Project Period (FY) 2014-04-01 – 2017-03-31
Project Status Completed (Fiscal Year 2016)
Budget Amount *help
¥3,640,000 (Direct Cost: ¥2,800,000、Indirect Cost: ¥840,000)
Fiscal Year 2016: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000)
Fiscal Year 2015: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2014: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
Keywordsニューラルネットワーク / 学習アルゴリズム / 勾配法 / 準ニュートン法 / 大規模データ / 勾配学習アルゴリズム / 凸化誤差関数 / 分割学習 / 分散学習
Outline of Final Research Achievements

In this research, it is a purpose to enable the approximation model by the feedforward neural networks for the function or the system with the highly nonlinear behavior and huge data by the following studies. Specifically, “Distribution of large scale data including highly nonlinear characteristics using statistical method”, and “Improvement of robustness of training algorithm by convexity of error function and its decentralization”. Aimed at the development of the proposed algorithm to solve the complexity and scale of the training problem was not feasible with conventional methods. Furthermore, this approach is useful for the circuit modeling for the design and optimization, where analytical formulas are not available or original model is computationally too expensive. A neural model is trained once, and can be used again and again. This avoids repetitive circuit simulations where a change in the physical dimension requires a re-simulation of the circuit structure.

Report

(4 results)
  • 2016 Annual Research Report   Final Research Report ( PDF )
  • 2015 Research-status Report
  • 2014 Research-status Report
  • Research Products

    (12 results)

All 2017 2016 2015 2014

All Journal Article (4 results) (of which Peer Reviewed: 4 results,  Acknowledgement Compliant: 2 results) Presentation (8 results)

  • [Journal Article] A Novel quasi-Newton-based Training using Nesterov's Accelerated Gradient for Neural Networks2016

    • Author(s)
      Hiroshi Ninomiya
    • Journal Title

      Proc. 2016 International Conference on Artificial Neural Networks (ICANN'16)

      Volume: - Pages: 540-540

    • Related Report
      2016 Annual Research Report
    • Peer Reviewed
  • [Journal Article] Neural Network Training based on quasi-Newton Method using Nesterov’s Accelerated Gradient2016

    • Author(s)
      Hiroshi Ninomiya
    • Journal Title

      Proc. IEEE TENCON 2016

      Volume: - Pages: 51-54

    • Related Report
      2016 Annual Research Report
    • Peer Reviewed / Acknowledgement Compliant
  • [Journal Article] Reconfigurable Dynamic Logic Circuit Generating t-Term Boolean Functions Based on Double-Gate CNTFETs2014

    • Author(s)
      Manabu Kobayashi, Hiroshi Ninomiya, Yasuyuki Miura and Shigeyoshi Watanabe
    • Journal Title

      IEICE Trans. on Fundamentals.

      Volume: E97-A Pages: 1051-1058

    • NAID

      130004770827

    • Related Report
      2014 Research-status Report
    • Peer Reviewed
  • [Journal Article] Distributed Robust Training of Multilayer Neural Networks Using Normalized Risk-Averting Error2014

    • Author(s)
      Hiroshi Ninomiya
    • Journal Title

      Proc. 2014 IEEE Symposium Series on Computational Intelligence (IEEE/SSCI'14 and IEEE/CCMB'14)

      Volume: 1 Pages: 134-140

    • Related Report
      2014 Research-status Report
    • Peer Reviewed / Acknowledgement Compliant
  • [Presentation] ネストロフの加速勾配を用いたQuickprop学習法の高速化2017

    • Author(s)
      Shahrzad Mahboubi,二宮 洋
    • Organizer
      2017年 電子情報通信学会 総合大会
    • Place of Presentation
      名古屋,名城大学
    • Year and Date
      2017-03-22
    • Related Report
      2016 Annual Research Report
  • [Presentation] 潜在クラスモデルのニューラルネットワークによる学習2016

    • Author(s)
      吉本昌史,小林 学,二宮 洋
    • Organizer
      2016年 電子情報通信学会 ソサイエティ大会
    • Place of Presentation
      北海道,北海道大学
    • Year and Date
      2016-09-20
    • Related Report
      2016 Annual Research Report
  • [Presentation] ネステロフの加速勾配を用いた準ニュートン学習法に関する研究2016

    • Author(s)
      三宅あかり,越森恵莉菜,二宮 洋
    • Organizer
      2016年 電子情報通信学会 ソサイエティ大会
    • Place of Presentation
      北海道,北海道大学
    • Year and Date
      2016-09-20
    • Related Report
      2016 Annual Research Report
  • [Presentation] ネステロフの加速準ニュートン法による学習アルゴリズムの提案2016

    • Author(s)
      二宮 洋
    • Organizer
      電子情報通信学会 信学技報 非線形問題研究会, vol.115, no.425, NLP2015-141
    • Place of Presentation
      九州工業大学
    • Year and Date
      2016-01-28
    • Related Report
      2015 Research-status Report
  • [Presentation] 再構成可能論理回路の設計法と各種方式の比較2016

    • Author(s)
      嘉藤淳紀,渡辺重佳,二宮 洋,小林 学,三浦康之
    • Organizer
      電子情報通信学会 信学技報, VLD2015-77, CPSY2015-109, RECONF2015-59
    • Place of Presentation
      慶應義塾大学
    • Year and Date
      2016-01-19
    • Related Report
      2015 Research-status Report
  • [Presentation] 粒子群最適化におけるローカル化及び寿命付きリーダーの有効性に関する研究2015

    • Author(s)
      佐伯 誠,坂下善彦,二宮 洋
    • Organizer
      情報処理学会 第77回 全国大会
    • Place of Presentation
      京都大学
    • Year and Date
      2015-03-17 – 2015-03-19
    • Related Report
      2014 Research-status Report
  • [Presentation] ダブルゲート型トランジスタを用いた再構成可能論理回路の設計法2014

    • Author(s)
      嘉藤淳紀,渡辺重佳,二宮 洋,小林 学,三浦康之
    • Organizer
      電子情報通信学会 信学技報, CPM2014-126, ICD2014-69
    • Place of Presentation
      別府国際コンベンションセンター
    • Year and Date
      2014-11-25 – 2014-11-29
    • Related Report
      2014 Research-status Report
  • [Presentation] ローカル粒子群最適化における寿命付リーダーの有効性2014

    • Author(s)
      佐伯 誠,坂下善彦,二宮 洋
    • Organizer
      電子情報通信学会 基礎・境界ソサイエティ大会
    • Place of Presentation
      徳島大学
    • Year and Date
      2014-09-23 – 2014-09-26
    • Related Report
      2014 Research-status Report

URL: 

Published: 2014-04-04   Modified: 2018-03-22  

Information User Guide FAQ News Terms of Use Attribution of KAKENHI

Powered by NII kakenhi