An analysis on the performance of silicon implementations of backpropagation algorithms for artificial neural networks

Author

Reyneri, Leonardo M. ; Filippi, Enrica

Author_Institution

Dipartimento di Elettronica, Politecnico di Torino, Italy

Volume

40

Issue

12

fYear

1991

fDate

12/1/1991 12:00:00 AM

Firstpage

1380

Lastpage

1389

Abstract

The effects of silicon implementation on the backpropagation learning rule in artificial neural systems are examined. The effects on learning performance of limited weight resolution, range limitations, and the steepness of the activation function are considered. A minimum resolution of about 20÷22 bits is generally required, but this figure can be reduced to about 14÷15 bits by properly choosing the learning parameter η which attains good performance in presence of limited resolution. This performance can be further improved by using a modified batch backpropagation rule. Theoretical analysis is compared with ad-hoc simulations and results are discussed in detail

Keywords

VLSI; artificial intelligence; learning systems; neural nets; Si; VLSI; activation function; artificial neural networks; backpropagation algorithms; learning rule; limited weight resolution; performance; range limitations; silicon implementations; simulations; steepness; Algorithm design and analysis; Artificial neural networks; Backpropagation algorithms; Circuits; Computational modeling; Computer networks; Multilayer perceptrons; Performance analysis; Silicon; Very large scale integration;

fLanguage

English

Journal_Title

Computers, IEEE Transactions on

Publisher

ieee

ISSN

0018-9340

Type

jour

DOI

10.1109/12.106223

Filename

106223