A kernel-based two-stage NU-support vector clustering algorithm

Support Vector Clustering is a kernel-based method that utilizes the kernel trick for data clustering. However it is only able to detect one cluster of non-convex shape in the feature space. In this study, we propose an alternative method using two-stage v-SVC to cluster data into several groups. The two-stage v-SVC is used to calculate the centroid of the sphere for each cluster in the feature space, and the K-means procedure is used to refine the clustering result iteratively. A mechanism is provided to control the position of the cluster centroid to work against outliers. Experimental results have shown that our method compares favorably with other kernel based clustering algorithms, such as KKM and KFCM, on several synthetic data sets and UCI real data.