运用组合加权距离函数的多指标面板数据聚类方法及应用  

The Method and the Application Multi-Index Panel Data Clustering Based on Combination Weighting Distance Function

作　　者：王少波 贾来喜[2] 

机构地区：[1]武警工程大学研究生管理大队,陕西西安710086 [2]武警工程大学军事基础教育学院,陕西西安710086

出　　处：《武警工程大学学报》2015年第6期1-4,共4页Journal of Engineering University of the Chinese People＇s Armed Police Force

摘　　要：针对多指标面板数据聚类分析的样品分类和空间地域划分，提出了一个多指标面板数据的综合聚类分析方法。从多元统计和层次分析角度，构造了多指标面板数据的样本距离函数。通过实证分析证明，该方法在能够满足系统分析的统一性要求的同时，保证指标之间的不相关；能够克服时间维度上均值处理的缺陷，信息损失较少；体现了指标间的主次差别。Based on multi-index panel data clustering analysis, we present a comprehensive cluster analysis method of multi-index panel data. From the viewpoint of multivariable statistical and analytic hierarchy process, the sample distance function of multi-index panel data is constructed. Empirical analysis shows this method can meet the need of the system analysis,and can overcome the shortcomings of the time dimension, the information loss is less, and the difference of the index is reflected.

关 键 词：多指标面板数据 聚类分析 样本距离函数 

分 类 号：G232.1[文化科学]