为了快速有效的检测聚类的边界点,提出基于网格核密度的自适应边界点检测算法ADAPT(An Adaptive Grid Kernel-Density-Based BoundaryPoints Detecting Algorithm for Spatial Database with Noise),使用网格核密度更精确地拟合网格在其邻域内的密度,采用自适应选取网格近邻策略更好地反应对象的空间分布特征.实验结果表明:该算法可以在含有任意形状、不同大小和不同密度的数据集上快速有效地检测出聚类的边界点.
In order to detect the boundary points of clusters effectively ,this paper presents an adaptive grid kernel-density-based boundarypoints detecting algorithm for spatial database with noise,ADAPT,which uses the concept of grid kernel density for the accuracy of grid density and a novel adaptive strategy for neighbor selection based on spatial object distribution. As shown by our experimental results, ADAPT detect boundary points effectively and efficiently on various datasets.