A Method of Generating Rules for a Kernel Fuzzy Classifier

A method of generating rules for a kernel fuzzy classifier is introduced. For this method, firstly, the initial sample space is mapped into a high dimensional feature space by selecting the appropriate kernel function. Then in the feature space, the proposed dynamic clustering algorithm dynamically separates the training samples into different clusters and finds out the support vectors of each cluster. For each cluster, a fuzzy rule is defined with ellipsoidal regions. Finally, the rules are adjusted by genetic algorithms. This classifier with such fuzzy rules is evaluated by two typical data sets. For this classifier, the learning time is short, the classification accuracy is better and the speed of classification is quick.