K均值算法使用的聚类准则函数的误差平方和准则，通过反复迭代优化聚类结果，使所有样本到各自所属类别的中心的距离平方和达到最小。-K-means clustering algorithm uses squared error criterion function and criteria through iterative optimization clustering result, all the samples to the respective classes of the center square of the distance and minimized.