A novel dynamic structural neural network with neuron-regeneration and neuron-degeneration mechanisms

  • Authors:

  • Yingtung Hsiao;Chenglong Chuang;Joeair Jiang;Chiang Wang;Chengchih Chien

  • Affiliations:

  • Department of Electrical Engineering, Tamkang University, Taipei, Taiwan, China;Department of Electrical Engineering, Tamkang University, Taipei, Taiwan, China;Department of Bio-Industrial Mechanical Engineering, National Taiwan University, Taipei, Taiwan, China;Department of Electrical Engineering, Tamkang University, Taipei, Taiwan, China;Department of Electrical Engineering, Tamkang University, Taipei, Taiwan, China

  • Venue:
  • ISNN'05 Proceedings of the Second international conference on Advances in Neural Networks - Volume Part I
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents a novel neural network model with free structure style called dynamic structural neural network (DSNN). The neurons are randomly generated in a virtual three-dimensional space. The structure of the DSNN is reconfigurable for increasing the learning capacity of the network. Therefore, in this work, a structure reconfiguration algorithm is also proposed to achieve this goal. Finally, several simple pattern recognition problems are applied to the proposed DSNN to demonstrate the efficiency and performance of the network.