Pattern classification using a self-organizing neural network

Abstract

A modified self-organizing neural network is presented. The network is based on a laterally inhibited neural network model developed by Kohonen. Preliminary work is focused on the application of the neural model to the classification of feature vectors extracted from texture images. Results obtained from the self-organizing neural classifier are compared with the results of a feedforward neural network trained with the back propagation algorithm as well as with those of a statistically based classifier.