In this paper, the problem of fast learning algorithm for multi-layered feedforward neural network (MLFNN) is discussed. A new fast backpropagation (FB-P) learning algorithm is proposed, By the analysis of FB-P learning algorithm, a modified FB-P (MFB-P) learning algorithm is presented. Simulations are run with the problem of XOR for B-P, FB-P and MFB-P, and the corresponding results indicate that MFB-P or FB-P converges much more quickly than B-P and MFB-P has much better generalization than FB-P or B-P.<>


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Fast learning algorithms for multi-layered feedforward neural network


    Contributors:
    Min Liang (author) / Shi-Xi Wang (author) / Yong-Hong Luo (author)


    Publication date :

    1994-01-01


    Size :

    296138 byte




    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Fast Learning Algorithms for Multi-Layered Feedforward Neural Network

    Liang, M. / Luo, Y.-H. / Wang, S.-X. et al. | British Library Conference Proceedings | 1994



    Reinforcement Learning Using Feedforward Neural Network with Memory Mechanism

    Ozawa, S. / Shiraga, N. | British Library Online Contents | 2003


    Incremental Learning Algorithm for Feedforward Neural Network with Long-Term Memory

    Kobayashi, M. / Ozawa, S. / Abe, S. | British Library Online Contents | 2002


    Feedforward Neural-Network Controller for Ship Steering

    Balasuriya, B. A. A. P. / Hoole, P. R. P. / International Federation of Automatic Control | British Library Conference Proceedings | 1995