CaltechAUTHORS
  A Caltech Library Service

A learning algorithm for multi-layer perceptrons with hard-limiting threshold units

Goodman, Rodney M. and Zeng, Zheng (1994) A learning algorithm for multi-layer perceptrons with hard-limiting threshold units. In: Proceedings of IEEE Workshop on Neural Networks for Signal Processing. IEEE , Piscataway, NJ, pp. 219-228. ISBN 0780320263. https://resolver.caltech.edu/CaltechAUTHORS:20190315-142359536

[img] PDF - Published Version
See Usage Policy.

465Kb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20190315-142359536

Abstract

We propose a novel learning algorithm to train networks with multilayer linear-threshold or hard-limiting units. The learning scheme is based on the standard backpropagation, but with "pseudo-gradient" descent, which uses the gradient of a sigmoid function as a heuristic hint in place of that of the hard-limiting function. A justification that the pseudo-gradient always points in the right down hill direction in error surface for networks with one hidden layer is provided. The advantages of such networks are that their internal representations in the hidden layers are clearly interpretable, and well-defined classification rules can be easily obtained, that calculations for classifications after training are very simple, and that they are easily implementable in hardware. Comparative experimental results on several benchmark problems using both the conventional backpropagation networks and our learning scheme for multilayer perceptrons are presented and analyzed.


Item Type:Book Section
Related URLs:
URLURL TypeDescription
https://doi.org/10.1109/nnsp.1994.366045DOIArticle
Additional Information:© 1994 IEEE. The research described in this paper was supported by ARPA under grants number AFOSR-90-0199 and N00014-92-5-1860.
Funders:
Funding AgencyGrant Number
Advanced Research Projects Agency (ARPA)UNSPECIFIED
Air Force Office of Scientific Research (AFOSR)AFOSR-90-0199
Office of Naval Research (ONR)N00014-92-J-1860
Record Number:CaltechAUTHORS:20190315-142359536
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20190315-142359536
Official Citation:R. M. Goodman and Zheng Zeng, "A learning algorithm for multi-layer perceptrons with hard-limiting threshold units," Proceedings of IEEE Workshop on Neural Networks for Signal Processing, Ermioni, Greece, 1994, pp. 219-228. doi: 10.1109/NNSP.1994.366045
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:93887
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:15 Mar 2019 21:53
Last Modified:03 Oct 2019 20:58

Repository Staff Only: item control page