Fast Training of a Fuzzy Classifier with Pyramidal Membership Functions

摘要

In this paper we discuss a fuzzy classifier with pyramidal membership functions and its training method. First we divide the training data for each class into several dusters and for each cluster we define a fuzzy rule with a hyperbox region that includes all the training data in the cluster and define a pyramidal membership function for the hyperbox. Then we tune the fuzzy rules, i.e., the locations of the hyperboxes and the slopes of the membership functions successively until there is no improvement in the recognition rate for the training data. We evaluate our method using two benchmark data sets and compare the performance with other classifiers.