中文说明：应用背景isodata聚类算法可以通过类的自动合并与分裂，得到较为合理的类型数目c。它的聚类效果相比于c-均值算法而言具有更好的聚类效果，实现过程更为复杂。适用于聚类各种场合。关键技术 isodata聚类算法中，起始聚合中心的选取对聚类过程和结果都有较大的影响，如果选择的好，则算法的收敛速度快，聚类质量高。相比较于c-均值算法而言，c-均值算法简单，isodata算法复杂，isodata算法的类型数目可变。
English Description：
Application backgroundISODATA clustering algorithm can automatically merge and split the class, and get the reasonable number of C. Its clustering effect is better than the c- means algorithm, and the realization process is more complex.. Suitable for clustering all kinds of occasions.Key TechnologyIn the ISODATA clustering algorithm, the selection of the initial aggregation center has a large effect on the clustering process and results. If the selection is good, the convergence speed of the algorithm is fast and the clustering quality is high.. Compared with the average c- algorithm, the c- means is simple, the complexity of ISODATA algorithm, and the number of ISODATA algorithm is variable..