对于模式识别分类，利用K均值聚类，对原始数据进行分类。比较易于收敛，十分好用-For pattern recognition classification using the K-means clustering, and classification of the original data. Relatively easy to convergence, very easy to use