Tuning of a fuzzy classifier by solving inequalities

In this paper, we develop a novel method for tuning the sensitivity parameters of membership functions used in fuzzy classifiers created from numerical data. In the proposed method, tuning is done by solving a set of inequalities. Therefore, we first derive a set of inequalities that the sensitivity parameters must satisfy to correctly classify the data used for tuning. Then we discuss a method to solve these inequalities from which the sensitivity parameters are obtained. We demonstrate the effectiveness of the method using two classification problems, namely, classification of commonly used iris data and recognition of vehicle licence plates. The results are compared with those obtained by using the existing tuning method and with those by neural networks