Modular Rough Fuzzy MLP: Evolutionary Design

Abstract

This article describes a way of designing a hybrid system for classification and rule generation, in soft computing paradigm, integrating rough set theory with a fuzzy MLP using an evolutionary algorithm. An l-class classification problem is split into l two-class problems. Crude subnetworks are initially obtained for each of these two-class problems via rough set theory. These subnetworks are then combined and the final network is evolved using a GA with restricted mutation operator which utilizes the knowledge of the modular structure already generated, for faster convergence. The GA tunes the fuzzification parameters, and the network weights and structure simultaneously, by optimizing a single fitness function.