Increasing Attraction of Pseudo-Inverse Autoassociative Networks

  • Authors:

  • Dmitry O. Gorodnichy;Alexandre M. Reznik

  • Affiliations:

  • Department of Computing Science, University of Alberta, T6G 2H1 Edmonton Alberta, Canada. E-mail: dmitri@cs.ualberta.ca;IPMMS, Cybernetics Center of Ukrainian Academy of Sciences, 252650 Kiev, Ukraine

  • Venue:
  • Neural Processing Letters
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

We show how partial reduction ofself-connections of the network designed with the pseudo-inverselearning rule increases the direct attraction radius of the network.Theoretical formula is obtained. Data obtained by simulation arepresented.