• DocumentCode
    416687
  • Title

    Universal learning networks with adaptive node functions

  • Author

    Goto, Kenichi ; Hirasawa, Kotaro ; Hu, Jinglu

  • Author_Institution
    Graduate Sch. of Information, Production & Syst., Waseda Univ., Fukuoka, Japan
  • Volume
    3
  • fYear
    2003
  • fDate
    4-6 Aug. 2003
  • Firstpage
    2415
  • Abstract
    In this paper, a new architecture of neural networks is proposed named universal learning networks with adaptive node functions (ULNs with ANF), which aims at improving the learning performance and generalization ability of neural networks in terms of constructing node functions especially fitted for a certain problem. The simulation results show that the proposed method is effective for constructing networks.
  • Keywords
    generalisation (artificial intelligence); learning (artificial intelligence); learning systems; neural net architecture; adaptive node functions; neural networks; universal learning networks;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    SICE 2003 Annual Conference
  • Conference_Location
    Fukui, Japan
  • Print_ISBN
    0-7803-8352-4
  • Type

    conf

  • Filename
    1323624