基于累积平均密度的聚类方法

Clustering algorithm based on cumulative average density

针对DBSCAN算法存在的参数敏感性和不能区分相连的不同密度的簇等缺陷,提出了一种基于DBSCAN算法的改进算法。算法提出了累积平均密度的概念,用来作为簇合并的依据,弱化了密度阈值Minpts的作用;选取密度最大的对象作为初始聚类中心,按照密度由高到低的顺序进行聚类,具有一定的层次性,因此支持变密度数据集聚类。最后,用数据集对算法进行了聚类实验。实验结果表明,改进算法具有一定的参数鲁棒性,对于相连的不同密度的簇,能够达到理想的聚类效果。

There exist two defects in the DBSCAN algorithm： input sensitivity, unable to distinguish clusters which have different density and are adjacent to one another. To solve these defects, an im- proved algorithm based on DBSCAN is proposed. The algorithm uses cumulative average density to de- terminate whether one threshold clustering Minpts , according to cluster can be merged with another or not, has weakened the role of density chooses the object with the maximal density as the beginning center object, does the density from high to low, which is hierarchical to a degree, and hence sup- ports clustering datasets with variable dens ments. The results show that the improved lty. algo In the end, datasets are used to do clustering experi- m has robustness of can achieve desired effect when clustering dataset with clusters of variab parameters to some extent, and le density linked together.