A note on linearly constrained Bayes estimator in elliptical models

摘要

<br/>A constrained Bayes estimator is established for the location parameter of an elliptically contoured distribution, when it is suspected that some linear constraints may hold on the parameter space. Under Jeffrey’s prior, the posterior distribution is proposed and the exact form of the constrained empirical Bayes estimator is derived for a general form of balanced loss function. Performance of the estimators is discussed using bootstrap simulation and a real data is analyzed to show the applicability of the proposal.

关键词