VGA-Classifier: Design and applications

A method for pattern classification using genetic algorithms (GAs) has been recently described in [1], where the class boundaries of a data set are ap proximated by a fixed number H of hyperplanes. As a consequence of fixing H a priori, the classifier suffered from the limitation of overfitting (or u nderfitting) the training data with an associated loss of its generalizatio n capability. In this paper, we propose a scheme for evolving the value of H automatically using the concept of variable length strings/chromosomes. T he crossover and mutation operators are newly defined in order to handle va riable string lengths. The fitness function ensures primarily the minimizat ion of the number of misclassified samples, and also the reduction of the n umber of hyperplanes. Based on an analogy between the classification princi ples of the genetic classifier and multilayer perceptron (with hard limitin g neurons), a method for automatically determining the architecture and the connection weights of the latter is described.