|
Activity Number:
|
77
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 2, 2009 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #303338 |
|
Title:
|
A Bayesian Approach to Population Parameter Orderings
|
|
Author(s):
|
Feiyi Jia*+ and Gary C. McDonald
|
|
Companies:
|
MED Institute and Oakland University
|
|
Address:
|
1 Geddes Way, West Lafayette, IN, 47906,
|
|
Keywords:
|
Ranking and selection ; Indifference zone ; Subset selection ; Order statistics ; Smallest credible subset
|
|
Abstract:
|
This article focuses on the orderings of population parameters, and attaches a probabilistic statement to each permutation of such orderings from a Bayesian point of view. Combined with suitable natural conjugate priors, discrete binomial data as well as continuous normal data are considered. The posterior distribution of the parameter permutations is obtained. Various ranking and selection problems can be addressed rather easily. Computer simulation can be used as a way of verifying the analytical derivation (and vice versa), or as an alternative when the analytical expression is not of a closed form.
|
