华东师范大学学报(自然科学版) ›› 2008, Vol. 2008 ›› Issue (5): 59-65.

• 应用数学,统计学 • 上一篇    下一篇

Buhlmann-Straub信度模型的估计与检验

唐国强1,2   

  1. 1.华东师范大学 统计系, 上海 200062;2.桂林工学院 数理系, 广西桂林 541004
  • 收稿日期:2007-07-15 修回日期:2008-04-07 出版日期:2008-09-25 发布日期:2008-09-25
  • 通讯作者: 唐国强

Estimating and testing parameter for Buhlmann-Straub model(Chinese)

TANG Guo-qiang1,2   

  1. 1.Department of Statistics, East China Normal University, Shanghai 200062,China;2. Department of Mathematics and Physics, Guilin University of Technology, Guilin Guangxi 541004, China
  • Received:2007-07-15 Revised:2008-04-07 Online:2008-09-25 Published:2008-09-25
  • Contact: TANG Guo-qiang

摘要: 研究了Buhlmann-Straub信度模型的参数估计和随机效应的检验,以及它们的统计性质. 参数估计采用两步估计法,用F检验对随机效应进行检验.在两步估计法中, 利用正交变换得到了总平均的估计, 然后采用拟合常数法得到了方差的估计,并证明了该估计是无偏估计.通过残差平方和构造了F检验统计量对是否有随机效应进行检验,并得出了检验的势函数是检验方差的增函数.最后将估计和检验的方法应用到实例中, 得到了较好的效果.

关键词: B-S信度模型, 正交变换, 拟合常数, 方差检验, B-S信度模型, 正交变换, 拟合常数, 方差检验

Abstract: In this Paper, the parameters and random effect of Buhlmann-Straub credibility model were researched.Two-stage estimate was used to estimate the parameter and F-test was used to test the random effect.In two-stage estimate, used orthogonal transformation to estimate aggregate average and fitting constants method to estimate variance which was unbiased.Used regression sum of squares to get the F-test statistic in order to test the random effect, the power of the test is increasing function of the test variance. AT last, the methods were used in a example that get good result.

Key words: orthogonal transformation, fitting constants method, test variance, Buhlmann-Straub model, orthogonal transformation, fitting constants method, test variance

中图分类号: