摘要
线性充分性概念由Ｂａｋｓａｌａｒｙ ａｎｄ ｋａｌａ［１］，Ｄｒｙｇａｓ［２，３］和Ｍｕｅｌｌｅｒ［４，５］引入广义线性模型Ｅ（Ｙ）＝Ｘβ，Ｃｏｖ（Ｙ）＝Ｖ≥０．本文拓广了这一概念 提出关于任一给定可估函数ｃ＇β的线性充分性概念，得到了线性充分性理论的一些进一步结果．本文还分析了观测值的一般线性变换对广义线性模型的影响，得到了由变换给可估函数的ＢＬＵＥ和ＢＬＵＥ之方差变化的一般结果.
This paper considers the general Gauss-Markoff model: E(Y) = Xβ, Cov (Y) = V≥ 0. The notion of linear sufficiency introduced by Baksalary and Kala[1], Drygas[2,3], Mueller[4,5]for estimating the set of all estimable functions is extended for estimating any specific estimable function. Some further contributions to the theory of linear sufficiency are provided. Moreover, the effects of transforming the observables Y into FY are analysed with respect to the BLUE and the covariance of the BLUE.
出处
《应用数学学报》
CSCD
北大核心
2001年第3期350-360,共11页
Acta Mathematicae Applicatae Sinica
