A Coreset-Based Semi-supverised Clustering Using One-Class Support Vector Machines

The traditional one-class support vector machines problem can be transformed into solving the minimum enclosing ball problem by the use of the corset. In this paper, the notion of the corset is applied to a semi-supervised clustering using one-class support vector machines. Experimental results show that this proposed algorithm not only can maintain the clustering performance, but also can decrease the running time of the clustering method.