Improving neural net convergence

M. Alder; S.G. Lim; P. Hadingham; Y. Attikiouzel

doi:10.1016/0167-8655(92)90030-4

Back

Journal article

Peer reviewed

Improving neural net convergence

M. Alder, S.G. Lim, P. Hadingham and Y. Attikiouzel

Pattern Recognition Letters, Vol.13(5), pp.331-338

1992

DOI: https://doi.org/10.1016/0167-8655(92)90030-4

Files and links (1)

url

Link to Published Version *Subscription may be requiredView

Abstract

This paper shows a certain equivalence between the 3-layer feed forward back-propagation neural net of Rumelhart et al. and the committee net of Nilsson. This is used to produce an improvement in the performance of the former. It is found that (a) the number of epochs taken is reduced by a factor of between 6 and 10, (b) the same taken is reduced by a factor of about 20, and (c) the net converges under conditions when the back-propagation algorithm is trapped in local minima and fails to converge.

Details

Title: Improving neural net convergence
Authors/Creators: M. Alder (Author/Creator) - The University of Western Australia
S.G. Lim (Author/Creator) - The University of Western Australia
P. Hadingham (Author/Creator) - The University of Western Australia
Y. Attikiouzel (Author/Creator) - The University of Western Australia
Publication Details: Pattern Recognition Letters, Vol.13(5), pp.331-338
Publisher: Elsevier B.V.
Identifiers: 991005542300007891
Murdoch Affiliation: Murdoch University
Language: English
Resource Type: Journal article

Metrics

52 Record Views

1 Times Cited - Web of Science

InCites Highlights

These are selected metrics from InCites Benchmarking & Analytics tool, related to this output

Citation topics: 4 Electrical Engineering, Electronics & Computer Science; 4.61 Artificial Intelligence & Machine Learning; 4.61.493 Neural-Fuzzy Integration
Web Of Science research areas: Computer Science, Artificial Intelligence
ESI research areas: Engineering