一般线性模型基于最大特征根的相对效率

The Relative Efficiency in General Linear Model Based on the Largest Relative Eigenvalue

定义了一般线性模型参数的最小二乘估计(LS估计)与BLU估计的一种新的相对效率,给出了当设计矩阵满秩,协方差阵非负定时这种相对效率的下界,最后讨论了其与广义相关系数(zρ3)之间的关系.

This paper defines a relative efficiency in singular G-M model based on the largest eigenvalue, Finds the lower bound of it and gives the relation between the general relative coefficient.