Title

Strategies for the Parallel Training of Simple Recurrent Neural Networks

Authors

Peter J. McCann, Washington University in St Louis
Barry L. Kalman, Washington University in St Louis

Document Type

Technical Report

Department

Computer Science and Engineering

Publication Date

1994-01-01

Filename

WUCS-94-15.PDF

DOI:

10.7936/K7QZ2865

Technical Report Number

WUCS-94-15

Abstract

Two concurrent implementations of the method of conjugate gradients for training Elman networks are discussed. The parallelism is obtained in the computation of the error gradient and the method is therefore applicable to any gradient descent training technique for this form of network. The experimental results were obtained on a Sun Sparc Center 2000 multiprocessor. The Sparc 2000 is a shared memory machine well suited to coarse-grained distributed computations, but the concurrency could be extended to other architectures as well.

Comments

Permanent URL: http://dx.doi.org/10.7936/K7QZ2865

Recommended Citation

McCann, Peter J. and Kalman, Barry L., "Strategies for the Parallel Training of Simple Recurrent Neural Networks" Report Number: WUCS-94-15 (1994). All Computer Science and Engineering Research.
https://openscholarship.wustl.edu/cse_research/337