摘要
传统的聚类分析方法利用数据对平均值的偏差，研究变量之间的相互关系。当它应用于区域化变量的分类时，因缺乏对区域化变量的空间结构的考虑，对合理解释区域化变量之间的相互关系不利。空间聚类分析方法比较好地解决了这个问题。本文介绍了该方法的基本原理及其算法。
Traditional cluster analysis,using the correlation coefficient or someother similar measurements,is commonly applied to evaluatethe correlationsamongcertain variables,But there are some dificulties when it is used to classify regionalizedvariables,Since it does not consider the spatial variations of the variahies it deals with.A new cluster analysis method,named Spatial Cluster Analysis,is presentd here.And the result of its application indicates that the spatial clsteranalysis is morereasonable than that of traditional cluster analysis in the decision of thecorrclation ofregionalized variables.
