Analysis of natural gradient descent for multilayer neural networks


Natural gradient descent (NGD) is an on-line algorithm for redefining the steepest descent direction. An analysis of NGD for training a multilayer neural network is presented using statistical mechanics. The performance can be significantly improved using NGD algorithm and can be used for both the transient and asymptotic stages of learning.

Divisions: College of Engineering & Physical Sciences > Systems analytics research institute (SARI)
Additional Information: ©1999 The American Physical Society
Publication ISSN: 1550-2376
Last Modified: 29 Nov 2023 10:33
Date Deposited: 01 Apr 2019 11:59
Full Text Link:
Related URLs: http://www.scop ... tnerID=8YFLogxK (Scopus URL)
https://journal ... hysRevE.59.4523 (Publisher URL)
PURE Output Type: Article
Published Date: 1999-04-01
Authors: Rattray, M.
Saad, D.



Version: Published Version

| Preview

Export / Share Citation


Additional statistics for this record