A minimax linear estimator for linear parameters under restrictions in form of inequalities
暂无分享,去创建一个
For a linear modely = Xβ + e with additional information about β given by β′ H β ≦ σ 2 a minimax linear estimation for a parameter γ =Cβ is derived for a quadratic loss. Therefore a matrix optimization problem is solved which can be use for several other minimax problems.
[1] C. R. Rao,et al. Linear Statistical Inference and its Applications , 1968 .
[2] O. Bunke,et al. Minimax linear, ridge and shrunken estimators for linear parameters , 1975 .
[3] S. Zacks. The theory of statistical inference , 1972 .