说明：  ：在有限区域内分布的稀疏不均的、具有一定分布结构的海量数据集的高效、高精度聚 类问题是一个尚未完全圆满解决的难题。针对Affinity Propagation 聚类算法(AP)的不足之处， 提出了两个改进型的聚类算法
(n limited areas of uneven distribution of sparse, has certain distribution structure of the mass datasets of high efficiency, high precision together Problem is a perfectly solved problem has not yet completely. For Affinity Propagation clustering algorithm (AP) of the deficiencies, Put forward the two improved version of the clustering algorithm)