文档介绍:: .
多指标面板数据聚类方法及应用
以行业一次能源消费
xtz/
:据为例
王双英王群伟3,曹泽4
(,FI照276826; ,南京210016;
,苏州215325; ,合肥230601)
摘要:综合考虑而板数裾多指标屮因变砧指标特征及其与自变砧指标的相关关系,通过定义因变量自协方差及自变量与因变量协方差以构建面板数据相似及相关性测度距离函数, 在引入自组织竞争网络算法的基础之上,提岀了多指标面板数据聚类的方法。以我国 1996-2008年44个行业煤炭、天然气、电力消耗量及国际石油价格面板数据进行实际应用, 验证了新方法聚类结果更为显著的优点。
关键字:自协方差;协方差;自组织竞争网络聚类分析;系统广义矩估计;一次能源消费
The Clustering Method Based on the Multi-dimensional Panel Data and its
Application
Wang Shuang-ying1,2, Wang Qun-wei2,3 , CaoZe2’
( of Management,Qufu Normal University, Rizhao 276826, China; 2. College of Economics and Management,Nanjing University of Aeronautics and Astronautics,Nanjing 210016, China; of Business,Soochow University,Suzhou 215021,China;
University of Architecture,Hefi 230601,China )
Abstract:Based on multivariable characteristics of panel data,by defining the autocovariance of the dependent variable and the covariance between the dependent and independent variables to constructing the similar and related distances,as introducing the petitive works in, it proposes one new clustering method. As an example,the method is used to analysis the energy consumption in 44 industries and the international oil price 1996-2008, and the results is prominent.
Key words: Autocovariance; Covariance; petiti