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 accepts input amount of K　then the object n data is divided into k cluster so that the obtained clustering meet: high similarity in the same cluster　while in different cluster object similarity is smaller. Cluster similarity is the use of the mean of each cluster obtained by objects in a　center　(center of gravity) to calculate.