Invariant Linearly Sufficient Transformations of the General Gauss-Markoff Model. Estimation and Testing
给出了使一般高斯-马尔可夫模型在线性充分变换下保持估计和检验信息所需的充分条件，并应用于缺失值理论。
In this paper the sufficient conditions for a linearly sufficient transformation of the general Gauss-Markoff model cy = X , D(y) = a2V preserving information needed for estimation X,B, a2 and an estimable parametric function X'3 and testing linear hypotheses, are given. The results are applied to the theory of missing values.