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。一般都采用均方差作为标准测度函数. k个聚类具有以下特点：各聚类本身尽可能的紧凑，而各聚类之间尽可能的分开。

该算法的最大优势在于简洁和快速。算法的关键在于初始中心的选择和距离公式。

-First, choose k objects　n data object as initial cluster centers　and for the rest of the other objects, according to their similarity (distance) These cluster centers, respectively assign them to its most similar ( cluster centers represent) clustering　and then calculate each cluster center obtained new cluster (the cluster mean all objects)　repeats this process until the beginning of the standard measurement function converges. Are generally used as the standard deviation measurement function k clusters has the following characteristics: Each cluster itself as compact as possible, but as much as possible to separate between the clusters. The biggest advantage of this algorithm is simple and fast. The key algorithm is the selection and initial center of the distance formula.