【摘要】 本文在一般增长曲线模型下研究其回归系数线性估计的可容许性。令Q={DUF:D为t×p阶常数矩阵,F为n×l阶常数矩阵},在矩阵损失函数（d-KBL）（d-KBL）′下给出了可估函数KBL的估计DYF在估计类Q₁={DYF:DYF∈Q,DA≠K,CF=L}中是可容许估计的充要条件;在DA=K,CF=L时DYF在估计类Q中是可容许估计的充要条件。由此还得出了一元的Gauss-Markov模型和多元模型在线性估计类中是可容许估计的充要条件。更多还原

【Abstract】 The admissibility of linear estimate of regression coefficients under the General Growth Curve Model is studied. Let Q={DYF: D and F are constant matrices of order t×p and n×1, respectively} and KBL is linear estimable, the necessary and sufficient conditions for DYF to be admissible estimate under loss function （d - KBL） （d - KBL）’ in class Qand in class Q while DA=K and CF=L are given, A corollary of the necessary and sufficient conditions for an estimable function to be admissible estimate in the whole linear estimate class under the Gauss-Markov model and multivariate linear model are obtained. Where Q₁={DYF: DYF∈Q, DA)K, CF= L}.更多还原