Unlabeled Data Classification via Support Vector Machines and k-means Clustering