A learning result for continuous-time recurrent neural networks

被引:12
|
作者
Sontag, ED [1 ]
机构
[1] Rutgers State Univ, Dept Math, Hill Ctr, Piscataway, NJ 08854 USA
关键词
recurrent neural networks; system identification; computational learning theory;
D O I
10.1016/S0167-6911(98)00006-1
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The following learning problem is considered, for continuous-time recurrent neural networks having sigmoidal activation functions. Given a "black box" representing an unknown system, measurements of output derivatives are collected, for a set of randomly generated inputs, and a network is used to approximate the observed behavior. It is shown that the number of inputs needed for reliable generalization (the sample complexity of the learning problem) is upper bounded by an expression that grows polynomially with the dimension of the network and logarithmically with the number of output derivatives being matched. (C) 1998 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:151 / 158
页数:8
相关论文
共 50 条
  • [1] A learning result for continuous-time recurrent neural networks
    Sontag, Eduardo D.
    Systems and Control Letters, 1998, 34 (03): : 151 - 158
  • [2] A Continuous-time Learning Rule for Memristor-based Recurrent Neural Networks
    Zoppo, Gianluca
    Marrone, Francesco
    Corinto, Fernando
    2019 26TH IEEE INTERNATIONAL CONFERENCE ON ELECTRONICS, CIRCUITS AND SYSTEMS (ICECS), 2019, : 494 - 497
  • [3] ON THE DYNAMICS OF SMALL CONTINUOUS-TIME RECURRENT NEURAL NETWORKS
    BEER, RD
    ADAPTIVE BEHAVIOR, 1995, 3 (04) : 469 - 509
  • [4] Noisy recurrent neural networks: The continuous-time case
    Das, S
    Olurotimi, O
    IEEE TRANSACTIONS ON NEURAL NETWORKS, 1998, 9 (05): : 913 - 936
  • [5] Complete controllability of continuous-time recurrent neural networks
    Sontag, E
    Sussmann, H
    SYSTEMS & CONTROL LETTERS, 1997, 30 (04) : 177 - 183
  • [6] Global dissipativity of continuous-time recurrent neural networks with time delay
    Liao, XX
    Wang, J
    PHYSICAL REVIEW E, 2003, 68 (01):
  • [7] Parameter space structure of continuous-time recurrent neural networks
    Beer, Randall D.
    NEURAL COMPUTATION, 2006, 18 (12) : 3009 - 3051
  • [8] Global convergence of continuous-time recurrent neural networks with delays
    Zhao, Weirui
    Zhang, Huanshui
    ADVANCES IN NEURAL NETWORKS - ISNN 2006, PT 1, 2006, 3971 : 109 - 114
  • [9] Vector Equalization based on Continuous-Time Recurrent Neural Networks
    Mostafa, Mohamad
    Teich, Werner G.
    Lindner, Juergen
    6TH INTERNATIONAL CONFERENCE ON SIGNAL PROCESSING AND COMMUNICATION SYSTEMS (ICSPCS'2012), 2012,
  • [10] Output convergence analysis of continuous-time recurrent neural networks
    Liu, DR
    Hu, SQ
    PROCEEDINGS OF THE 2003 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS, VOL III: GENERAL & NONLINEAR CIRCUITS AND SYSTEMS, 2003, : 466 - 469