A Study on Memory Storing Abilities in Neural Networks with Stimulus-Response Scheme

• KATADA Naofumi

  Studies of Information Science, Hyogo University of Education

• NISHIMURA Haruhiko

  Studies of Information Science, Hyogo University of Education

Bibliographic Information

Other Title

• 刺激-応答機構に基づくニューラルネットワークの動的学習能

Abstract

We investigate the processes of dynamic learning under the stimulus-response scheme and examine the abilities of memory storing, comparing the chaotic neural network with the conventional Hopfield network. It is found that the memory-engraving capacities are much better than those of the Hopfield network. The relation to the Lyapunov exponent suggests that the learning ability is activated through the chaotic (complex) dynamics.

Journal

• IEICE technical report. Neurocomputing

  IEICE technical report. Neurocomputing 99 (494), 15-22, 1999-12-11

  The Institute of Electronics, Information and Communication Engineers

Keywords

• chaos
• stochastic
• stimulus-response
• neural network
• learning
• plasticity

Details

• CRID: 1573950402236973696
• NII Article ID: 110003234384
• NII Book ID: AN10091178
• Text Lang: ja
• Data Source:
  • CiNii Articles