Pattern classification using genetic algorithms: Determination of H

A methodology based on the concept of a variable string length GA (VGA) is developed for determining automatically the number of hyperplanes for model ing the class boundaries in a GA-classifier. The genetic operators and fitn ess function are defined to take care of the variability in chromosome leng th. It is proved that the method is able to arrive at the optimal number of misclassifications after a sufficiently large number of iterations, and wi ll need a minimal number of hyperplanes for this purpose. Experimental resu lts on different artificial and real life data sets demonstrate that the cl assifier. using the concept of a variable length chromosome, can automatica lly determine an appropriate value of the number of hyperplanes, and also p rovide performance better than that of the fixed length version. Its compar ison with another approach using a VGA is provided. (C) 1998 Elsevier Scien ce B.V. All rights reserved.