K-means for clustering analysis