Universal approximation by phase series and fixed-weight networks

  • Authors:

  • Neil E. Cotter;Peter R. Conwell

  • Affiliations:

  • Electrical Engineering Department, University of Utah, Salt Lake City, UT 84112 USA;Electrical Engineering Department, University of Utah, Salt Lake City, UT 84112 USA

  • Venue:
  • Neural Computation
  • Year:
  • 1993

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this note we show that weak (specified energy bound) universal approximation by neural networks is possible if variable synaptic weights are brought in as network inputs rather than being embedded in a network. We illustrate this idea with a Fourier series network that we transform into what we call a phase series network. The transformation only increases the number of neurons by a factor of two.