A note on james-stein and bayes empiricl bayes estimators |
| |
Authors: | Tze Fen Li |
| |
Institution: | Department of Mathematics , University of North Carolina at Charlotte , Charlotte, North charlotte, 28223 |
| |
Abstract: | In an empirical Bayes decision problem, a simple class of estimators is constructed that dominate the James-Stein estimator, A prior distribution A is placed on a restricted (normal) class G of priors to produce a Bayes empirical Bayes estimator, The Bayes empirical Bayes estimator is smooth, admissible, and asymptotically optimal. For certain A rate of convergence to minimum Bayes risk is 0(n-1)uniformly on G. The results of a Monte Carlo study are presented to demonstrate the favorable risk bebhavior of the Bayes estimator In comparison with other competitors including the James-Stein estimator. |
| |
Keywords: | admissible asymptotically optimal bayes empirical Bayes empirical Bayes smooth |
|
|