Some efficient simple rules in Γ-minimax estimation
Brani Vidaković, David Rı́os Insua · Communications in Statistics - Simulation and Computation · 1994
We discuss the r-minimax estimation of normal means under squared-error loss. Since this problem is computationally intensive, we study linear r-minimax estimation. We identify contexts in which linear rules are adequately efficient, in terms of the ratio of their risk to the optimal risk. We also provide strategies that may work in other contexts.