说明：  最大最小距离法是模式识别中一种基于试探的类聚算法，它以欧式距离为基础，取尽可能远的对象作为聚类中心。因此可以避免K-means法初值选取时可能出现的聚类种子过于临近的情况，它不仅能智能确定初试聚类种子的个数，而且提高了划分初试数据集的效率。 该算法以欧氏距离为基础，首先初始一个样本对象作为第1个聚类中心，再选择一个与第1个聚类中心最远的样本作为第2个聚类中心，然后确定其他的聚类中心，直到无新的聚类中心产生。最后将样本按最小距离原则归入最近的类。
(The maximum minimum distance method is a trial based clustering algorithm in pattern recognition. It is based on Euclidean distance and takes the object as far as possible as the clustering center. It can not only intelligently determine the number of clustering seeds in the initial test, but also improve the efficiency of dividing the initial test data set. The algorithm is based on Euclidean distance. Firstly, a sample object is taken as the first cluster center, and then a sample farthest from the first cluster center is selected as the second cluster center. Then other cluster centers are determined until no new cluster center is generated. Finally, the samples are classified into the nearest class according to the minimum distance principle.)