Abstract:
Clustering algorithm is a very important research direction in data mining. So far, lots of clustering algorithms adapted to the large-scale and high-dimension data base have been proposed. The density-based algorithm is one of the typical researching directions. On the basis of CABDET algorithm, this paper presents a new algorithm called GFCABD. The new algorithm puts the theory of grid into the density-tree based clustering algorithm, thus improves the efficiency of clustering effectively and reduces the cost of I/O.
Key words:
Clustering,
Density,
Grid,
Density-tree
摘要: 聚类算法是数据挖掘领域中一个非常重要的研究方向。人们已经提出了许多适用于大规模的、高维的数据库的聚类算法。基于密度的聚类算法是其中一个比较典型的研究方向。该文以CABDET算法为基础,提出了一种基于密度树的网格快速聚类算法,该算法将网格的原理运用到基于密度树的聚类算法中,有效地提高了聚类的效率,降低了I/O的开销。
关键词:
聚类,
密度,
网格,
密度树
CLC Number:
LAI Jianzhang; NI Zhiwei;LIU Zhiwei. A Grid Fast Clustering Algorithm Based on Density-tree[J]. Computer Engineering, 2006, 32(17): 69-70,8.
赖建章;倪志伟;刘志伟. 一种基于密度树的网格快速聚类算法的研究[J]. 计算机工程, 2006, 32(17): 69-70,8.