Fuzzy SVM Based on Triangular Fuzzy Numbers
Qiang Zhi He, Congxin Wu, Eric C.C. Tsang · 2007
Support vector machine (SVM) is novel type learning machine, based on statistical learning theory, whose tasks involve classification, regression or novelty detection. Traditional SVM classifies the data with numerical features. However, in most cases of real world, there are much more data with fuzzy features. It is difficult to apply traditional SVM to fuzzy data directly to classify. In this paper, we provide a fuzzy SVM for the data with triangular fuzzy number features. The designing fundamentals and method of computation and realization are given. The experiment results show that the new method proposed in this paper is more effective and practical. This new method optimizes the classified result of support vector machine and enhances the intelligent level of support vector machine.