一种基于位置距离和属性特征结合的聚类方法 |
| |
引用本文: | 石亚冰,黄予. 一种基于位置距离和属性特征结合的聚类方法[J]. 人天科学研究, 2013, 0(3): 51-54 |
| |
作者姓名: | 石亚冰 黄予 |
| |
作者单位: | 广西师范学院计算机与信息工程学院,广西南宁530001 |
| |
摘 要: | 利用空间坐标和属性特征的有机结合,定义了3种曼哈顿空间距离,用matlab编程给出了基于该空间距离的ACA—Cluster聚类算法,并对山东省生态环境质量进行了聚类分析和类型分区。实验表明,该方法可以较好地反映出空间位置邻近和属性特征相似的空间聚类要求。
|
关 键 词: | 空间距离 空间对象 ACA—Cluster聚类 空间聚类 空间数据挖掘 |
The Method of Combinate the Coordinate and Attributes for Clustering |
| |
Abstract: | Based on combination of coordinate and attribute , three Manhattan distance based spatial measurements are defined. And the computation of ACA-Cluster on the basis of spatial distance is proposed by matlab programming. Used this method, the environment quality of Shandong Province is put into clustering analysis and type division. It is made clear that the spatial clustering can express both positional neighborhood and attributive similarity after our experimens. |
| |
Keywords: | Spatial Distance Spatial Object ACA-Cluster |
本文献已被 维普 等数据库收录! |