Nearly Exponential Approximation for Neural Networks

Abstract

In this paper we prove that any real function in L_P (C) defined on a compact and convex subset of R^dcan be approximated by a sigmoidal neural network with one hidden layer , that we call nearly exponential.