The superiority of Bayes estimators of the estimable function of regression coefficient matrix and the covariance matrix in multivariate linear model
-
摘要: 本文研究了在设计阵非列满秩情况下多元线性模型的Bayes估计问题. 假定回归系数矩阵和协方差阵具有正态--逆Wishart先验分布,运用Bayes理论导出了回归系数矩阵的可估函数和协方差阵的同时Bayes估计. 然后在Bayes Mean Square Error (BMSE)准则和Bayes Mean SquareError Matrix (BMSEM)准则下,证明了可估函数和协方差阵的Bayes估计优于广义最小二乘(GeneralizedLeast Square, GLS)估计. 另外, 在Bayes Pitman Closeness (BPC)准则下研究了可估函数的Bayes估计的优良性. 最后, 进行了Monte Carlo模拟研究, 进一步验证了理论结果.
-
关键词:
- 可估函数 /
- 正态--逆Wishart先验 /
- BMSE准则 /
- BMSEM准则 /
- BPC准则
Abstract: In this paper, the parameter estimation problem in a multivariate linear model is investigated when the design matrix is non-full rank, the joint prior of regression coefficient matrix and covariance matrix is assumed to be the normal-inverse Wishart distribution. By using the Bayes theory, the Bayes estimation of estimable function of regression coefficient matrix and covariance matrix are derived. Then we prove that the Bayes estimation of estimable function and covariance matrix are superior to the corresponding generalized least square (GLS) estimators under the criteria of Bayes mean square error (BMSE) and Bayes mean square error matrix (BMSEM). In addition, under the Bayes Pitman Closeness (BPC) criterion, the superiority of the Bayes estimation of estimable function is also investigated. Finally, a Monte Carlo simulation is carried out to verify the theoretical results.-
Key words:
- estimable function /
- normal-inverse Wishart prior /
- BMSE criterion /
- BMSEM criterion /
- BPC criterion
-
表 1 λ12和λ34的模拟值
Tab. 1 Simulation values of λ12 and λ34
q = 6 q = 7 q = 8 q = 9 q = 10 Ψ1 λ12 2.027 1.975 2.038 2.075 2.032 λ34 3.211 4.478 4.818 4.724 5.331 Ψ2 λ12 2.041 2.042 2.004 2.055 2.018 λ34 4.679 4.260 4.784 4.142 5.688 
下载: 导出CSV
-
[1] RAO C R. Linear Statistical Inference and Its Applications[M]. New York: Wiley, 1973. [2] 陈敏, 韦来生. 线性模型中回归系数的可估函数和误差方差的同时Bayes估计的优良性[J]. 应用数学学报, 2014, 37(5): 865-877. http://www.cnki.com.cn/Article/CJFDTOTAL-YYSU201405009.htm [3] 王松桂, 史建红, 尹素菊, 等. 线性模型引论[M]. 北京: 科学出版社, 2004. [4] BOX G E P, TIAO G C. Bayesian Inference in Statistical Analysis[M]. Massachusetts: Addison-Wesley, 1973. [5] XU K, HE D J. The superiority of Bayes estimators in a multivariate linear model with respect to normal-inverse Wishart prior[J].Acta Mathematicae Applicatae Sinica, English Series, 2015, 31(6): 1003-1014. doi: 10.1007/s10114-015-3649-2 [6] ZHANG W P, WEI L S, CHEN Y. The superioritities of Bayes linear unbiased estimator in multivariate linear models[J]. Acta Mathematicae Applicatae Sinica, English Series, 2012, 28(2): 383-394. doi: 10.1007/s10255-012-0150-x [7] ANDERSON T W. An Introduction to Multivariate Statistical Analysis[M]. New Jersey: John Wiley and Sons, 2003. [8] 许凯, 何道江, 徐兴忠. 正态--逆Wishart先验下多元线性模型中经验Bayes估计的优良性[J]. 数学年刊, 2014, 35A(3): 267-284. http://www.cnki.com.cn/Article/CJFDTOTAL-SXNZ201403002.htm [9] ZHANG X D. Matrix Analysis and Application[M]. Beijing: Tsinghua University Press, 2004. [10] LIU J S. An Introduction to Wishart Distributions[M]. Beijing: Science Press, 2005. [11] PITMAN E J G. The closest estimates of statistical parameters[J]. Proc Camb Phil Soc, 1937, 33: 212-222. doi: 10.1017/S0305004100019563 [12] RAO C R. Some comments on the minimum mean square as criterion of estimation statistics and related topics[M]//Statistics and Related Topics. Amsterdam: North-Holland Publishing Co, 1981: 123-143. [13] RAO C R, KEATING J P, MASON R L. The Pitman nearness criterion and determination[J]. Communications in Statistics--Theory and Methods, 1986, 15: 3173-3191. doi: 10.1080/03610928608829302 [14] KEATING J P, MASON R L, SEN P K. Pitman's measure of closeness: A comparison of statistical estimators[M]//Society for Industrial and Applied Mathematics. Philadelphia: [s.n.], 1993: 571-581. [15] WEI L S, ZHANG W P. The superiorities of Bayes linear minimun risk estimation in linear mode[J]. Communications in Statistics-Theory and Methods, 2007, 36: 917-926. doi: 10.1080/03610920601036333 [16] 陈玲, 韦来生. 线性模型中回归系数和误差方差的同时,Bayes,估计的优良性[J]. 数学年刊, 2011, 32A(6): 763-774. http://www.cnki.com.cn/Article/CJFDTOTAL-SXNZ201106016.htm -
