主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2005, Vol. ›› Issue (5): 126-129.

• 论文 • 上一篇    下一篇

多因素数据重心预测方法及应用研究

张群, 张积林   

  1. 北京科技大学管理学院, 北京, 100083
  • 收稿日期:2004-11-16 修回日期:2005-08-15 出版日期:2005-10-28 发布日期:2012-03-07

Study of Multi-Factor Data Barycenter Forecasting and Applications

ZHANG Qun, ZHANG Ji-lin   

  1. Management School, University of Science and Technology of Beijing, Beijing 100083, China
  • Received:2004-11-16 Revised:2005-08-15 Online:2005-10-28 Published:2012-03-07

摘要: 本文基于数据重心概念,通过大量的研究和推导,提出数据重心参数估计理论,并利用数据重心法估计多项式回归预测模型的参数,应用于我国钢材消费量的预测。从应用的结果看,增加了我国目前钢材预测的方法。本方法能最大限度地平滑预测模型的误差,而且应用条件比最小二乘法宽松;也不会因为个别残差较大的异常点而对预测结果产生不稳定性,从而提高了拟合和预测的稳健度,计算更简捷。

关键词: 数据重心法, 参数估计, 钢材预测, 稳健法

Abstract: A new and useful parameter estimation method was presented in light of the stability of gravity center of object in this paper.This method could deal with the fitting and forecasting of economy volume and could greatly decrease the errors of the fitting and forecasting results.Moreover,the strict hypothetical conditions in least squares method were not mecessary in the method presented in this paper,which overcame the shortcomings of least squares method and expanded the application of data barycentre method.Application to the steel consumption volume forecasting was presented in this paper.It is shown that the result of fitting and forecasting was satisfactory.From the comparison between data barycentre forecasting method and least squares method,we could conclude that the fitting and forecasting results using data barycentre method were more stable than that using least squares regression forecasting method,and the computation of data barycentre forecasting method was simpler than that of least squares method.As a result,the data barycentre method was convenient to use in technical economy.

Key words: data barycentre method, parameter estimation, steel forecasting

中图分类号: