带有约束的增长曲线模型中回归系数线性估计的可容许性与泛容许性(英文)  

ADMISSIBILITY AND GENERAL ADMISSIBILITY OF LINEAR ESTIMATES OF REGRESSION COEFFICIENTS IN GROWTH CURVE MODEL WITH RESPECT TO A RESTRICTED PARAMETER SET

作　　者：胡雪梅[1,2] 王志忠[1] 高骥忠[3] 

机构地区：[1]中南大学数学科学与计算技术学院,湖南长沙410075 [2]重庆工商大学数学与统计学院,重庆400067 [3]南华工商学院计算机系,广东广州510720

出　　处：《数学杂志》2010年第1期20-34,共15页Journal of Mathematics

基　　金：Supported by theNatural Science of Hunan(03JJY3065);Doctor's Start-up Research Foundation of Chongqing Technology and Business University(08-52204)

摘　　要：本文研究了参数受约束的增长曲线模型中多元回归系数线性估计的可容许性和泛容许性.利用线性估计类中的八种最优标准和φ函数,得到了在三个等价子类中线性估计可容许以及回归系数线性估计泛容许的充要条件.本文的结论推广了覃红等人的工作.In the article, we investigate the admissibility and general admissibility of linear estimates of multivariate regression coefficients in the growth curve model with respect to a restricted parameter set. By using eight kinds of optimality criteria among linear estimates and φ function, we obtain the sufficient and necessary conditions that linear estimates among three identical subclasses are admissible and that linear estimates of regression coefficients are general admissible. The conclusions generalize the work of Qin et al.

关 键 词：容许性 泛容许性 增长曲线模型 

分 类 号：O212.4[理学—概率论与数理统计]