Type-Ⅱ Fuzzy Support Vector Machine

Fuzzy support vector machine, which is an extension of support vector machine, cansolve the classification problems effectively in fuzzy environment. However, it is difficult todeal with classification problems in type-Ⅱ fuzzy environment. As a generalization of fuzzyset theory, type-Ⅱ fuzzy set theory is more effective for the delineation of existent fuzziness.For this purpose, this paper discusses support vector machine based on type-Ⅱ fuzzysamples which extend real samples to type-Ⅱ fuzzy samples. In order to deal with theproblems above, type-Ⅱ fuzzy samples are reduced to ordinary fuzzy samples via the meanreduction method. The credibility of fuzzy event is introduced on the basis of possibilitytheory, the definition of strong fuzzy linear separable for type-Ⅱ fuzzy samples is given, andthe type-Ⅱ fuzzy support vector machine based on input or output type-Ⅱ fuzzy samples isestablished. Example result shows the efficiency of the type-Ⅱ fuzzy support vectormachine.