基于相对密度的不确定数据聚类算法

Relative Density-based Clustering Algorithm over Uncertain Data

传统的基于相对密度的聚类算法有效地解决了密度聚类算法对参数敏感以及不能区分不同密度等级簇的问题。基于相对密度的不确定聚类算法,借用了相对密度算法的思想,根据不确定数据的特征,定义了不确定数据的距离公式、相对密度、核心点、密度可达等相关概念,从而提出了一种能够有效地处理不确定数据的新算法。数据仿真结果表明了该算法的有效性和可用性。

Traditional relative density-based clustering algorithm has advantage in handling shortcomings of user-defined parameters＇ sensitivity and distinguishing different hierarchy of density. This paper provided a new uncertain data clustering algorithm based on relative density, which defines distance formula, density ratio, core points and densityreachable,and can efficiently handle uncertain data. The simulation results illustrate the validity and availability of the algorithm.