Growing layers of perceptrons: introducing the Extentron algorithm

Concepts based on two observations of perceptrons are presented. When the perceptron learning algorithm cycles among hyperplanes, the hyperplanes may be compared in order to select one that gives a best split of the examples, and it is always possible for the perceptron to build a hyperplane that separates at least one example from all the rest. The authors describe the Extentron, which grows multi-layer networks capable of distinguishing nonlinearly separable data using the simple perceptron rule for linear threshold units. The resulting algorithm is simple, very fast, scales well to large problems, retains the convergence properties of the perceptron, and can be completely specified using only two parameters. Results are presented comparing the Extentron to other neural network paradigms and to symbolic learning systems