L2-SVM based fuzzy classifier with automatic model selection and fuzzy rule ranking

In this paper, a new scheme for constructing parsimonious fuzzy classifiers is proposed based on the L2-support vector machine (L2- SVM) technique, in which the model selection is automatically performed in terms of the radius-margin bound of the SVM and fuzzy rules are optimally generated from data. In order to identify the most influential induced fuzzy rules and remove the redundant ones, two novel indices for fuzzy rule ranking are proposed in this paper based on the L2-SVM learning results. They are named as α -values and ω -values of fuzzy rules in this paper. The experimental results on a high-dimensional benchmark problem have shown that by using the proposed indices the most influential fuzzy rules can be effectively selected, leading to a parsimonious fuzzy classifier which achieves better generalization performance than the well-known algorithms in the literature.