这里的k-means聚类，是事先给出原始数据所含的类数，然后将含有相似特征的数据聚为一个类中。-Where k-means clustering, is given in advance the number of classes contained in the original data, then the data contained similar characteristics grouped into a single class.