基于视觉原理的密度聚类算法的改进

Enhanced visual-based density clustering algorithm

结合基于视觉原理的密度聚类算法对初始化参数不敏感、能发现任意形状的聚类、能够找出最优聚类及一趟聚类算法快速高效的特点,研究可以处理混合属性的高效聚类算法.首先简单改进基于视觉原理的密度聚类算法,使之可以处理含分类属性的数据,进而提出一种两阶段聚类算法。第一阶段使用一趟聚类算法对数据集进行初始划分,第二阶段利用基于视觉原理的密度聚类算法归并初始划分而得到最终聚类。在真实数据集和人造数据集上的实验结果表明,提出的两阶段聚类算法是有效可行的。

The visual-based density clustering algorithm which is insensitive to initialized parameters,identify the data with any shape and can find the optimal cluster.The one-pass clustering algorithm is efficient and fast.Based on their features research was done on a clustering algorithm which can process the data with mixed attributes.At first,the visual-based density clustering algorithm was slightly improved,which enabled it to process data with categorical attributes.Then,the two-stage clustering algorithm was put forward.In the first stage,the single pass clustering algorithm was used to group the data as an original partition.In the second stage,the improved visual-based density clustering algorithm was used to merge the original partition so that the clusters were finally obtained.Experimental results of both actual and synthetic datasets show that the presented clustering algorithm is effective and practicable.