An improved k-means clustering algorithm based on dissimilarity | IEEE Conference Publication | IEEE Xplore