Unification of Fuzzy SVMs and Rule Extraction Methods through imprecise Domain Knowledge
Christian Moewes, Rudolf Kruse · 2008
In this paper, we want to moti-vate the combination of kernel-based methods with fuzzy rule extraction methods to describe uncertain do-mains by fuzzy models. We thus introduce and motivate the concept of a fuzzy support vector machine (FSVM) to incorporate imprecise-ness into kernel machines. Further-more, we present the idea of a posi-tive definite fuzzy classifier (PDFC), the rules of which are obtained by kernel-based models. We conclude with two vague conceptions to asso-ciate FSVM with PDFC to finally obtain understandable and meaning-ful fuzzy rules.