Rough support vector clustering

In this paper a novel kernel-based soft clustering method is proposed. This method incorporates rough set theoretic flavour in support vector clustering paradigm to achieve soft clustering. Empirical studies show that this method can find soft clusters having arbitrary shapes.