A Speedy Perceptron Algorithm Based on Distance

Abstract

In linear space, the classical perceptron algorithm is simple and practical. But when concerning the nonlinear space it is severely confined mainly on its signal layer structure. This paper analyzes the geometry characteristic of solve region in the pattern set, and presents a new algorithm based on the solve region. The new algorithm could find the better solve vector in the solve region on condition that the pattern space could distinguish in linear separable space, Otherwise, it could indicate that the solve region inexistence. And then, in the nonlinear separable space, the new algorithm starts with the distance of different categories in the pattern space, fits the curve according to cubic parametric spline curve firstly, and then fits discrimination function in a specific way on the basis of the specific problem. It is scientifically proved that the algorithm is feasible and effective. It solves the convergence in all the traditions, and finally enhances the speed of calculation.