Because of the shortcomings of the traditional K-means algorithm which is sensitive to select the initial clustering centers and easy converges to local optimization, the paper proposes An Clustering ...