北京航空航天大学学报(社会科学版) ›› 2010, Vol. 23 ›› Issue (3): 66-71.

• 论文 • 上一篇    下一篇

三维立体数据表的动态趋势聚类研究 ——以中国现代服务业为实例

孙晓丹   

  1. 北京航空航天大学 经济管理学院, 北京 100191
  • 收稿日期:2008-11-27 出版日期:2010-05-25 发布日期:2010-06-29
  • 作者简介:孙晓丹(1980-),男,黑龙江齐齐哈尔人,博士研究生,研究方向为多元统计与复杂数据分 析.
  • 基金资助:

    国家自然科学基金资助项目(70771004);国家自然科学基金创新研究群体科学基金资助项目(70521001)

Dynamic Trend Clustering Research Based on the Three-way Data

SUN Xiao-dan   

  1. School of Economics and Management, Beijing University of Aeronautics and Astronautics, Beijing 100191, China
  • Received:2008-11-27 Online:2010-05-25 Published:2010-06-29

摘要:

在信息技术快速发展的今天,数据形式的多样性使得对问题和现象的研究不再局限于单纯利用截面数据或时间序列数据进行分析。文章所研究的是一种由截面数据和时间序列数据共同 组成的具有立体数据结构的三维立体数据的分类问题,以定义的常数型均值以及动态非相似度为基础对其进行聚类分析。同时,以中国现代服务业的实际数据进行相应的实例验证 。结果证明,该种方法能有效的从规模和趋势上对三维立体数据进行分类。

关键词: 三维立体数据, 常数型均值, 动态非相似度, 聚类分析

Abstract:

Along with the rapid development of information and technology, the diversity an d numerousness of data forms have become a common problem confronted by data col lectors and researchers. The various researches not only analyzed and discussed problems by using the three-way data, but also in multi-angles and multi-dime nsions. The analysis object of this paper is the classification of three-way da ta, which is three-dimensional data composed of cross-sectional data and time series data. The paper tries to do clustering analysis based on constant mean an d dynamic dissimilarity, and uses modern service industry data of our country to make an empirical analysis.

Key words: three-way data, constant mean, dynamic dissimilarity, clustering analysis

中图分类号: