An Experimental Investigation of Gladun’s Theory of Automatic Classification

Ninth Midwest Artificial intelligence and Cognitive Science Conference

Volume

Ninth Midwest Artificial intelligence and Cognitive Science Conference

Contents

Gladun’s Pyramidal Net Classification Theory is a novel theory of automatic classification based on the idea of a hierarchical system of concepts formed through competitive concept formation procedures based on neural recognition metaphors. This paper addresses various observed phenomena of Gladun’s methods by describing and analyzing the application of the methods to a number of classification problems. The problems include the prediction of the values of a parabola, circle, and sphere.