Clustering is an important part of data mining. The existing clustering algorithm failed in the data set with uneven density distribution. In this paper, we propose a novel clustering algorithm relative density-based clustering algorithm for identifying diverse density clusters effectively called IDDC. It can effectively identify clusters in data sets with different densities and can also handle outliers. We first compute relative density for each data point. Then, the density peak points are screened and the initial clusters are obtained according to these peak points. The strategy for assigning the remaining points is to find unallocated points from the perspective of the cluster, which can effectively identify different density. In experiments, we compare the proposed algorithm IDDC with some existing algorithms on synthetic and real-world data sets. The results show that IDDC performs better than those existing algorithms, especially clustering on data set with uneven density distribution.

集群是数据挖掘的重要组成部分。现有的聚类算法在密度分布不​​均的数据集中失败。在本文中，我们提出了一种新的基于相对密度的聚类算法，该算法可以有效地识别各种密度聚类，称为IDDC。它可以有效地识别具有不同密度的数据集中的聚类，还可以处理离群值。我们首先计算每个数据点的相对密度。然后，筛选出密度峰值点，并根据这些峰值点获得初始聚类。分配剩余点的策略是从群集的角度查找未分配的点，这样可以有效地识别不同的密度。在实验中 我们将拟议的IDDC算法与一些现有的合成和现实数据集算法进行了比较。结果表明，IDDC的性能优于现有算法，尤其是在密度分布不​​均匀的数据集上的聚类。