线性模型中估计的可容许性
ADMISSIBILITY OF ESTIMATOR IN LINEAR MODEL
-
摘要: 本文推广了La Motte有关线性模型中可容许估计的结果,由此解决了一种方式分组随机效应模型中方差分量二次及非负二次估计的可容许问题,并证明了可容许估计类是一最小完备类。Abstract: In this paper, La Motte’s results about admissibility in linear model are generalized. By using these results, the admissibility of quadratio and non-negative quadratio estimation of variance components in random one-way model is solved. It is also proved that the admissible class is a minimum complete class.