Search

Clustering analysis is an unsupervised learning method, which has applications across many fields such as pattern recognition, machine learning, information security, and image segmentation. The density-based method, as one of the various clustering algorithms, has achieved good performance. However, it works poor in dealing with multidensity and complex-shaped datasets. Moreover, the result of this method depends heavily on the parameters we input. Thus, we propose a novel clustering algorithm (called the MST-DC) in this paper, which is based on the density core. Firstly, we employ the reverse nearest neighbors to extract core objects. Secondly, we use the minimum spanning tree algorithm to cluster the core objects. Finally, the remaining objects are assigned to the cluster to which their nearest core object belongs. The experimental results on several synthetic and real-world datasets show the superiority of the MST-DC to Kmeans, DBSCAN, DPC, DCore, SNNDPC, and LDP-MST. [ABSTRACT FROM AUTHOR]

Cluster analysis has been widely used in pattern recognition, image segmentation, document clustering, intrusion detection, market research, and so on. There have been many clustering algorithms for cluster analysis now. However, most of them are unsuitable for complex patterns with large variations in density and manifold structure. Also, they are not robust to noises. To overcome the above deficiencies, we propose a novel clustering algorithm (called DCLGMS) based on the gravity-mass-square ratio (LGMS) and density core (DCore) with a dynamic denoising radius. Our algorithm can obtain the correct clustering result in arbitrarily shaped datasets excluding noises without parameter settings. In this algorithm, DCore can maintain the shapes of clusters, and LGMS can help to extract local core points. In addition, a dynamic denoising radius is beneficial to detect noise points caused by arbitrary reasons and avoid interference at the beginning of the algorithm. The results of experiments on both synthetic datasets and real datasets show that our algorithm has excellent performance. [ABSTRACT FROM AUTHOR]