An extended class of minimax generalized Bayes estimators of regression coefficients

摘要

We derive minimax generalized Bayes estimators of regression coefficients inthe general linear model with spherically symmetric errors under invariantquadratic loss for the case of unknown scale. The class of estimatorsgeneralizes the class considered in Maruyama and Strawderman (2005) to includenon-monotone shrinkage functions.