本算法在vc++6.0中进行实验。分别就分解聚类和C-均值聚类两种方法在IRIS数据集上进行操作。分类前先将数据集中的样本顺序打乱形成混合数据。分解聚类中，采用前100个样本用对分法编制程序将数据分为两类。C-均值聚类采用全部的150个样本，将类别参数K设为3，将数据分为三类。-The algorithm in vc++6.0 in the experiment. Separate cluster and decomposition of two C-means clustering methods operate on the IRIS data set. Classification of samples before the first data set in order disrupt the formation of mixed data. Decomposition clustering using the top 100 on the sub-samples prepared with the program data into two categories. C-means clustering using all 150 samples, the type of parameter K set to 3, the data into three categories.