基于网格的密度树算法.pdfVIP

Proceedings of the Fourth International Conference on Machine Learning and Cybernetics, Guangzhou, 18-21 August 2005 A CLUSTERING ALGORITHM BASED ON BUILDING A DENSITY-TREE 1 1 1 2 WEI-DI DAI , YUE-XIAN HOU , PI-LIAN HE , XIAO-SHEN ZHENG 1Department of Computer Science and Technology, Tianjin University, Tianjin 300072, China 2College of Marine Science and Engineering, Tianjin University of Science and Technology, Tianjin 300457,China E-MAIL: davidy@126.com, yxhou@, plhe@, zhengxi aoshen@163.com Abstract: form all data points. Comparing with other clustering A new kind of clustering algorithm called CABDET is algorithms, DENCLUE generalizes partition-based methods presented in this paper. CABDET creates a tree structure for and hierarchical methods, and has good clustering every cluster, from which the neighbor’s radius of the current properties for data sets with large amounts of noise. object is calculated by the local density of its father node. However, the method is sensitive to the input parameter and Those unprocessed objects in the neighbor of the current noise threshold. The selection of such parameters may object are added to extend the tree structure until no new object is founded. Each density-tree is regarded as one cluster. significantly influence the quality of the clustering results. CABDET requires only one input parameter as the initial DBSCAN can quickly discover dif