提出基于密度的网格聚类算法GDcA,发现大规模空间数据库中任意形状的聚类.该算法首先将数据空间划分成若干体积相同的单元,然后对单元进行聚类只有密度不小于给定阈值的单元才得到扩展,从而大大降低了时间复杂性在GDcA的基础上,给出增量式聚类算法IGDcA,适用于数据的批量更新.

Although many clustering algorithms have been proposed so far, seldom was focused on high-dimensional and incremental databases. This paper introduces a grid density-based clustering algorithm GDCA. which discovers clusters with arbitrary shape in spatial databases. It first partitions the data space into a number of units, and then deals with units instead of points. Only those units with the density no less than a given minimum density threshold are useful in extending clusters. An incremental clustering algorithm----IGDCA is also presented, applicable in periodically incremental environment.

Supported by the National Natural Science Foundation of China under Grant No.69983011(国家自然科学基金);the NationalGrand Fundamental Research 973 Program of China under Grant No. G I999035807(国家重点基础研究发展规划973资助项目)