Automatic Parameter Tuning of Support Vector Clustering

Clustering is an unsupervised technique to group the data according to their mutual similarities. In this research paper we have introduced an improved version of Support Vector Clustering (SVC) technique with automatic parameter tuning. Using Gaussian kernel data points are mapped into a higher dimensional feature space. In Gaussian kernel we look for the minimal enclosing sphere. This sphere, when mapped back to the data space, separates into several components with irregular boundaries, enclosing separate clusters of points. There are two tuning parameters in SVC, soft margin penalty constant and width of the Gaussian kernel which are varied and used to attain smooth cluster boundaries. It is difficult to accurately select both of these parameters manually (by k-fold Cross Validation). So, we have employed an automatic RBF sigma parameter tuning technique which was originally proposed by Cheng Hsuan Li et.al for Support Vector Machines (SVM). It finds out the optimal kernel RBF sigma value for SVC, which is more efficient and accurate than k-fold cross validation method.