Learning in the recurrent random neural network

  • Authors:
  • Erol Gelenbe

  • Affiliations:
  • -

  • Venue:
  • Neural Computation
  • Year:
  • 1993

Quantified Score

Hi-index 0.00

Visualization

Abstract

The capacity to learn from examples is one of the most desirablefeatures of neural network models. We present a learning algorithmfor the recurrent random network model (Gelenbe 1989, 1990) usinggradient descent of a quadratic error function. The analyticalproperties of the model lead to a "backpropagation" type algorithmthat requires the solution of a system of n linear andn nonlinear equations each time the n-neuron network"learns" a new input-output pair.