Title :
Pattern classification using polynomial networks
Author :
Patrikar, A. ; Provence, J.
Author_Institution :
Dept. of Electr. Eng., Southern Methodist Univ., Dallas, TX, USA
fDate :
6/4/1992 12:00:00 AM
Abstract :
The use of polynomial networks is proposed for efficient implementation of the polynomial classifiers. The polynomial network is a trainable device consisting of a large number of simple processing elements. In this respect it is similar to artificial neural networks. A learning algorithm for the polynomial network is developed and its performance is compared with that of multilayer perceptrons.
Keywords :
learning systems; neural nets; pattern recognition; polynomials; artificial neural networks; learning algorithm; multilayer perceptrons; pattern classification; polynomial classifiers; polynomial networks; trainable device;
Journal_Title :
Electronics Letters
DOI :
10.1049/el:19920700
