Interval set representations of 1-v-r support vector machine multi-classifiers

Support vector machines (SVMs) are designed for linearly separating binary classes. Researchers have suggested various approaches, such as the one-versus-rest (1-v-r), one-versus-one (1-v-1) and DAGSVM, for applying SVMs to multi-classification problems. The 1-v-r approach tends to have a large training time, while the 1-v-1 and DAGSVM approaches often store a large number of SVMs. We have recently shown how traditional SVMs can be represented using interval or rough sets. In this paper, we extend the interval set formulation of SVMs to classifications that involve more than two classes that are separated using the 1-v-r approach. Our approach possesses several salient features. The presented work is especially useful for soft margin classifiers. Our approach seeks a balance by reducing the training time while storing fewer rules. Finally, our technique provides a semantic interpretation of the classification process, as opposed to the black-box SVM methods.