A Fuzzy Twin Support Vector Machine Algorithm

Although twin support vector machine (TSVM) has faster speed than traditional support vector machine for classification problem, it does not take into account the importance of the training samples on the learning of the decision hyper-plane with respect to the classification task. In this paper, fuzzy twin support vector machine (FTSVM) is proposed where a fuzzy membership value is assigned to each training sample. Here, training samples are classified by assigning them to the nearest one of two nonparallel planes that are close to their respective classes. Moreover, this method only requires solving a smaller size SVM-type problem as compared to SVMs where the classifier is obtained by solving a quadratic programming problem. Experiments on several UCI benchmark datasets show that FTSVM is effective and feasible compared with twin support vector machine(TSVM), fuzzy support vector machine(FSVM) and support vector machine(SVM).