k-means 算法接受输入量 k ；然后将n个数据对象划分为 k个聚类以便使得所获得的聚类满足：同一聚类中的对象相似度较高；而不同聚类中的对象相似度较小。聚类相似度是利用各聚类中对象的均值所获得一个“中心对象”（引力中心）来进行计算的。-K- means algorithm accept input k　Then could be divided into k n data object clustering in order to make the clustering obtained satisfy: clustering objects in the same similarity is higher　Unlike smaller clustering objects in the similarity. Cluster similarity is the use of the cluster of objects in the average obtained a "central object" (center of gravity) to calculated.