A TWO-LEVEL COEVOLUTIONARY APPROACH TO MULTIDIMENSIONAL PATTERN CLASSIFICATION PROBLEMS

Abstract

This paper proposes a coevolutionary classification method to discover classifiers for multidimensional pattern classification problems with continuous input variables. The classification problems may be decomposed into two subproblems, which are feature selection and classifier adaptation. A coevolutionary classification method is designed by coordinating the two sub-problems, whose performances are affected by each other. The proposed method establishes a group of partial sub-regions, defined by regional variable set, and then fits a finite number of classifiers to the data pattern by combining a genetic algorithm and a local adaptation algorithm in every sub-region. A cycle of the cooperation loop is completed by evolving the subregions based on the evaluation results of the fitted classifiers located in the corresponding sub-regions. The classifier system has been tested with well-known data sets from the UCI machine-learning database, showing superior performance to other methods such as the nearest neighbor, decision tree, and neural networks.