Multilayer Feedforward Networks with Non-Polynomial Activation Functions Can Approximate Any Function

17 Pages Posted: 31 Oct 2008

See all articles by Moshe Leshno

Moshe Leshno

Independent

Shimon Schocken

affiliation not provided to SSRN

Date Written: September 1991

Abstract

Several researchers characterized the activation functions under which multilayer feedforwardnetworks can act as universal approximators. We show that all the characterizationsthat were reported thus far in the literature ark special cases of the following general result:a standard multilayer feedforward network can approximate any continuous functionto any degree of accuracy if and only if the network's activation functions are not polynomial.We also emphasize the important role of the threshold, asserting that without it thelast theorem doesn't hold.

Keywords: Multilayer feedforward networks, Activation functions, role of threshold, Universal approximation capabilities, LP(μ) approximation

Suggested Citation

Leshno, Moshe and Schocken, Shimon, Multilayer Feedforward Networks with Non-Polynomial Activation Functions Can Approximate Any Function (September 1991). NYU Stern School of Business Research Paper Series, Forthcoming, Available at SSRN: https://ssrn.com/abstract=1289056

Shimon Schocken

affiliation not provided to SSRN

No Address Available

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
177
Abstract Views
1,455
Rank
306,387
PlumX Metrics