Minimax estimation for certain independent component districutions under welghted squared error loss

Abstract

Let X be a vactor of independent components with mean vectoro. θ We assume that the distribution of the j^th component is of the form i.e.a.varliant mixture of noramal distribution are minimax for the problem of estimating the vector under the loss function

Keywords:

• Location parameters
• decision theory
• James Stein estimaion