Characterization Of Admissibility For Linear Estimator And Its Superiority In General Linear Model

The theory about the admissibility in the Gauss-Markov model is comparative maturity, and includes integrated systemic results. In this paper, we study characterizations of admissible in the general linear model (Y, Xβ,ε|ε～(0,σ~2∑). We demonstrate that an admissible linear estimator is as the conditional generalized ridge-type estimation in the no constraint, equality constraint, inequality constraint general linear model. We study the superiority of this conditional generalized ridge-type estimation, and prove that it is superior to the restricted best linear unbiased estimator in terms of mean squares. We also give the choice of the matrix K.We first state general linear model, ridge-type estimator and general ridge-type estimator, and the constraint biased estimator. And then, we introduce some basic theory about matrix and some conclusion about the admissibility of estimator in Gauss-Markov model. In the third chapter, we discussion several equivalent characterization of the best linear unbiased estimation, we proved that admissible characterization of admissible of linear estimation is as conditional general ridge-type estimation in general linear model. A necessary and sufficient condition that homogeneous linear estimator is admissible estimator is obtained. In the fourth chapter, we discussion the characterization of admissible in the general linear model under equality constraint and inequality constraint, we give the necessary and sufficient condition that homogeneous linear estimator is admissible estimator, and by using the relationship between homogeneous and inhomogeneous linear estimator, we obtain the characterization of admissible inhomogeneous linear estimator. In the fifth chapter, we propose the conditional generalized ridge-type estimator of regression coefficient in restricted linear regression model, and prove that it is superior to the restricted best linear unbiased estimator in terms of mean squares error and mean squares error matrix, we also give the choice of parameters matrix K.