Genetic Learning of Fuzzy Rule-Based Classification Systems Cooperating with Fuzzy Reasoning Methods

In this paper, we present a multi-stage genetic learning process for obtaining linguistic Fuzzy Rule-Based Classification Systems, that integrates Fuzzy Reasoning Methods cooperating with the fuzzy rule base and learns the best set of linguistic hedges for the linguistic variable terms. We show the application of the genetic learning process to two well known sample bases, and compare the results with those obtained from different learning algorithms. The results show the good behaviour of the proposed method maintaining the linguistic description of the fuzzy rules. 1 Introduction In a supervised inductive learning process, a pattern classification system is developed by means of a set of classified examples used to establish a correspondence between the new pattern descriptions and the class labels. This system is a description of the concepts learnt and it can be expressed as a logic expression, a set of fuzzy rules, a decision tree or a neural network, among others. When the class..