On-line learning in multilayer neural networks

Abstract

We present an analytic solution to the problem of on-line gradient-descent learning for two-layer neural networks with an arbitrary number of hidden units in both teacher and student networks. The technique, demonstrated here for the case of adaptive input-to-hidden weights, becomes exact as the dimensionality of the input space increases.

Publication DOI: https://doi.org/10.1007/978-1-4615-6099-9_53
Divisions: College of Engineering & Physical Sciences > Systems analytics research institute (SARI)
Aston University (General)
Additional Information: © Springer Science+Business Media New York 1997
Uncontrolled Keywords: algorithm,design,measurement,performance,theory,verification
ISBN: 0-7923-99331
Last Modified: 02 Dec 2024 09:11
Date Deposited: 08 Jul 2009 10:36
Full Text Link:
Related URLs: https://link.sp ... -4615-6099-9_53 (Publisher URL)
PURE Output Type: Chapter
Published Date: 1997
Authors: Saad, David (ORCID Profile 0000-0001-9821-2623)
Solla, Sara A.

Download

Export / Share Citation


Statistics

Additional statistics for this record