统计研究 ›› 2010, Vol. 27 ›› Issue (9): 73-77.

• 论文 • 上一篇    下一篇

面板数据聚类方法及应用

李因果 何晓群   

  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2010-09-15 发布日期:2010-09-15

Panel Data Clustering Method and Application

Li Yinguo & He Xiaoqun   

  • Received:1900-01-01 Revised:1900-01-01 Online:2010-09-15 Published:2010-09-15

摘要: 基于面板数据的时序特征和截面特征,综合考虑面板数据“绝对指标”,“增量指标”及其“时序波动”特征,在重构面板数据相似性测度的距离函数和Ward聚类算法的基础上,提出了面板数据聚类方法。并以2003-2007年财政金融面板数据为例,对中国14个沿海开放城市进行了聚类分析,显示了良好的应用性。

关键词: 面板数据, 绝对指标, 增量指标, 变异指标, 聚类分析

Abstract: Based on time series and cross section’s characteristics of panel data. Under comprehensive consideration of the panel data’s characteristics with “absolute index”, “incremental index” and “time fluctuation”, this text reconstructs the distance function of similarity measure and Ward clustering algorithm, presents a panel data clustering analysis method. As an example, the method is used on analysis the financial panel data of 14 coastal cities of China from 2003 to 2007 and the clustering analysis results shows good applicability.

Key words: Panel, Data, Absolute, Index, Incremental, Index, Variation, Index, Clustering, Analysis