A bayes sequential procesure for selecting the most probable multinomial event
James Tompkins Ramey, Khursheed Alam · Communication in Statistics- Theory and Methods · 1980
The classical approach to the problem of ranking and selection is a sort of minimax approach. It is interesting to consider the Bayesian approach if some prior information is available with regard to the parameters of the underlying distributions. This paper deals with a Bayes sequential sampling procedure for selecting the most probable event from a multinomial distribution whose parameters are distributed a priri according to a Dirichlet distribution. The given rule is compared with other sampling rules which nave been considered in the literature.