k-means 算法的工作过程说明如下：首先从n个数据对象任意选择 k 个对象作为初始聚类中心；而对于所剩下其它对象，则根据它们与这些聚类中心的相似度（距离），分别将它们分配给与其最相似的（聚类中心所代表的）聚类；然后再计算每个所获新聚类的聚类中心（该聚类中所有对象的均值）；不断重复这一过程直到标准测度函数开始收敛为止。-k-means algorithm process as follows: First of all, the object data from the n choose k object as initial cluster centers　and the remaining for the other object, then according to their cluster center with those of the similarity (distance) respectively assigned to them with the most similar (represented by cluster center) clustering　obtained and then calculated for each new cluster center clustering (all objects in the cluster mean)　repeated this process until the standard measure of function until the beginning of convergence.