Abstract
In this paper some of the problems that arise in learning for feed-forward neural networks using backward error propagation (B.E.P.) are considered — notably rate of convergence and the size of the training set. Suggestions are made as to how improved gradient methods and the use of autoassociative networks together with cluster analysis techniques may address these difficulties.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Bibliography
Ackley, Hinton and Sejnowski “A Learning Algorithm for Boltzmann Machines”, Cognitive Sc. 9, 147–169, (1985).
Rumelhart, Hinton and Williams, “Learning Representations by Back Propagating Errors”, Nature, 533–536 323, (1986).
Bank, R.E. and Rose, D.J., “Global Approximate Newton Methods”, Numerische Mathematik, Springer Verlag, (1981).
Hinton, G., “Learning Translation Invarient Recognition.”, Preprint (1987).
Elman, J.L. and Zipser, D., “Learning the Hidden Structure of Speech”, ICS Report 8701, (1987).
Gorman, R.P. and Sejnowski, T.J., “Analysis of Hidden Layers in a Layered Network.” Neural Networks Vol 1, 75–90, (1988).
Johnson, S.C., “Hierarchical Clustering Schemes”, Psychometrika 32, 241254, (1967).
Bourlard, H., and Kemp, Y., “Auto-Association by Multilayer Perceptrons and Singular Value Decomposition”, Biol. Cybern. 59 291–294, (1988)
Hopfeld,.J.J., “Neural networks and physical systems with emergent collective computational properties”, Proc. Natl. Acad. Sci. USA, 79, 2554–2558, (1982).
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1990 British Computer Society
About this paper
Cite this paper
Kinsella, J.A. (1990). Training Neural Networks: Strategies and Tactics. In: Smeaton, A.F., McDermott, G. (eds) AI and Cognitive Science ’89. Workshops in Computing. Springer, London. https://doi.org/10.1007/978-1-4471-3164-9_13
Download citation
DOI: https://doi.org/10.1007/978-1-4471-3164-9_13
Publisher Name: Springer, London
Print ISBN: 978-3-540-19608-2
Online ISBN: 978-1-4471-3164-9
eBook Packages: Springer Book Archive