On the mapping capacity of feed forward neural networks

Gang WEI*, Sam KWONG, K. F. MAN

*Corresponding author for this work

Research output: Book Chapters | Papers in Conference ProceedingsConference paper (refereed)Researchpeer-review

Abstract

This paper addresses the mapping capacity of feed forward neural networks (FFNNs). We have proposed a new condition of the activation function for FFNNs to be capable of approximating any continuous maps defined on Rn. Based the result, we have constructed several FFNN based nonlinear predictors for speech signals, and studied their performances.

Original languageEnglish
Title of host publicationProceedings of the 1995 IEEE International Conference on Systems, Man and Cybernetics
PublisherIEEE
Pages468-471
Number of pages4
Volume1
ISBN (Print)0780325591
DOIs
Publication statusPublished - 1995
Externally publishedYes
Event1995 IEEE International Conference on Systems, Man and Cybernetics - Vancouver, Canada
Duration: 22 Oct 199525 Oct 1995

Publication series

NameProceedings of the IEEE International Conference on Systems, Man and Cybernetics
ISSN (Print)0884-3627

Conference

Conference1995 IEEE International Conference on Systems, Man and Cybernetics
Country/TerritoryCanada
CityVancouver
Period22/10/9525/10/95

Fingerprint

Dive into the research topics of 'On the mapping capacity of feed forward neural networks'. Together they form a unique fingerprint.

Cite this