基于密度的最小生成树聚类算法研究

基于密度的方法是一种相当有效的聚类方法,能够发现任意形状的聚类,对噪声数据不敏感,但是聚类结果严重依赖于用户参数的合理选择。针对其存在的问题,将最小生成树理论与基于密度的方法相结合,提出了一种基于密度的最小生成树聚类算法。通过构造、分割最小生成树得到确定样本空间划分的最小生成子树;根据子树特性,产生局部密度参数;并对生成子树进行局部密度聚类。理论分析和应用结果表明。该算法不仅体现了基于密度聚类方法的优点,聚类结果不依赖于用户参数的选择,使数据聚类更合理,特别是对大型数据库非常有效;也体现了数据分区的思想,使其可以并行执行,进一步提高了信息处理的时空效率和性能。

Research of Density-Based MST Clustering Algorithm

Density-based method is one of the effective and efficient clustering methods that can discover clusters with arbitrary shape and is insensitive to noise data.But some parameters should be determined by hand.However,it is very difficuh to determine them manually without any prior domain knowledge.To solve this problem,an applicable and efficient density-based minimum spanning tree clustering algorithm is presented.It aims at avoiding any parameter to be determined by hand ,and at the same time ,improving the efficiency of clustering.Theory analysis and experimental results show that this approach not only keeps the good features of DBSCAN,but also requires no input parameters and can make paralleled processing.