A Constructing Method of Fuzzy Classifier Using Kernel K-Means Clustering Algorithm

A constructing method of fuzzy classifier using kernel k-means clustering algorithm is introduced in this paper. This constructing method are divided into three phases, namely clustering phase, fuzzy rule created phase and parameters modified phase. Firstly, the original sample space is mapped into a high dimensional feature space by selecting appropriate kernel function. In the feature space, training samples are grouped into some clusters by kernel k means clustering algorithm. Then for each created cluster, a fuzzy rule is defined with the appropriate membership function. Finally, some parameters of fuzzy classifier are chosen by GAs. The experiment results show the proposed fuzzy classifier has very high classification accuracy by the comparison results with the similar approach, and has the better applied values.