Limiting the Risk of the Bayes Estimator for the Binomial Distribution | ||||
The Egyptian Statistical Journal | ||||
Article 1, Volume 35, Issue 2, December 1991, Page 191-199 PDF (2.37 MB) | ||||
Document Type: Original Article | ||||
DOI: 10.21608/esju.1991.315005 | ||||
![]() | ||||
Author | ||||
Reda Ibrahim Mazloum* | ||||
Faculty of Economics & Political Science, Cairo University, Egypt | ||||
Abstract | ||||
This article considers the Bayesian problem of estimating the probability of success θ of a binominal distribution when θ itself has a beta prior. The usual Bayes estimator for this situation has high risk if θ is far from the mode of the prior distribution. We suggest a rule which does not have this bad property and still perform well against the beta prior. This rule is a compromise between the Bayes rule and the MLE. | ||||
Keywords | ||||
Bayes Estimator; Beta Prior; Binomial Distribution; Maximum Likelihood Estimation; Risk Reduction | ||||
Statistics Article View: 148 PDF Download: 13 |
||||