Abstract
A Bayesian single sample solution to the problem of selecting the largest of k normal means is given. The solution appears robust with respect to variations in the prior distribution and parameters of the model.
Get full access to this article
View all access options for this article.