Geometric Rates of Approximation by Neural Networks

Abstract

Sets of functions for which errors in approximation by neural networks converge to zero geometrically fast are described.

Categories and tags

  • Konferenční příspěvek (mezinárodní konf.)