Magnified gradient function in adaptive learning : The MGFPROP algorithm

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review

1 Scopus Citations
Scopus Metrics
View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Original languageEnglish
Pages (from-to)42-43
Journal / PublicationElectronics Letters
Volume37
Issue number1
Publication statusPublished - 4 Jan 2001

Link(s)

DOI

DOI
Check@CityULib
Link to Scopus https://www.scopus.com/record/display.uri?eid=2-s2.0-0035120935&origin=recordpage
Permanent Link https://scholars.cityu.edu.hk/en/publications/publication(1dcdbad2-b42c-420b-91ff-6abf64770f15).html

Abstract

A new algorithm is proposed to solve the `flat spot' problem in back-propagation neural networks by magnifying the gradient function. Simulation results show that, in terms of the convergence rate and the percentage of global convergence, the new algorithm consistently outperforms other traditional methods.

Fingerprint

Publication fingerprints text

100
Algorithms Computer Science
100
Adaptive Learning Computer Science
100
algorithms INIS
100
learning INIS
100
Algorithm Mathematics
100
Gradient Function Keyphrases
100
Neural Network Chemical Engineering
66
convergence INIS
50
Convergence Rate Keyphrases
50
Global Convergence Keyphrases
Full Fingerprint ›

Citation Format(s)

[ RIS ] [ BibTeX ]
Magnified gradient function in adaptive learning: The MGFPROP algorithm. / Ng, Sin-Chun; Cheung, Chi-Chung; Leung, Shu-Hung.
In: Electronics Letters, Vol. 37, No. 1, 04.01.2001, p. 42-43.

Research output: Journal Publications and Reviews › RGC 21 - Publication in refereed journal › peer-review