Gauss-Markov估计关于误差分布的稳健性
Robustness of Gauss-Markov Estimator in Terms of Error Distributions
-
摘要: 对于一般线性模型y=X\beta+\varepsilon, 本文讨论了在广义均方误差准则及均方误差矩阵准则下, 未知参数\beta的可估函数X\beta的Gauss-Markov估计关于误差分布的稳健性, 分别给出了误差项\varepsilon的最大分布类, 使得误差项\varepsilon的分布在此范围内变动时, Gauss-Markov估计在相应准则下是最优估计Abstract: Robustness of Gauss-Markov estimator of estimable function of unknown parameter in terms of error distributions is discussed in general linear model. We explore the maximal distribution classes of error term, where Gauss-Markov estimator held its optimality by generalized mean squared errors criterion and by mean square error matrix criterion respectively
下载: