基于密度的空间聚类算法研究

基于密度的聚类算法作为数据挖掘方法中的一种主要方法,不仅可以从数据集中发现任意形状的簇,而且可以观察到一个并发的、完整的聚类结构,以及具有对噪声数据不敏感的特点。针对目前常用的几种基于密度的聚类算法及改进算法进行讨论,分析了这些密度聚类算法各自的优缺点,并且以地理信息系统为应用研究背景,提出了基于密度的聚类算法与GIS相结合,通过对多维数据属性特征的提取,扩展到多维数据的处理,在三维空间地形数据中的分析中取得了高效的聚类结果。

Research of Spatial Clustering Algorithms Based on Density

Spatial clustering algorithm based on density can find the cluster of random shape,and some subsequent and integrate clustering structure,and lack of sensitivity to noise.Expound several density based spatial clustering and some upswing method,and analyse the advantage and disadvantage of them.Then on background of GIS propose a method of combined density based spatial clustering algorithm and GIS,which extract the attribute of the multidimensional data can expand to handle the multidimensional data,and get the good effect from the analysis of the three dimension space.