Combining Selection and Estimation in the Search for the Largest BinomialParameter

摘要

For k > or = independent binomial populations, from which X sub; approx. = B (nitheta sub;) i = 1,...,k, have been observed, the problem of selecting the population with the largest theta-value and simultaneously estimating the theta-parameter of the selected population is considered. Under several loss functions, Bayes decision rules are derived and studied for independent Beta-priors. A fixed sample size look ahead procedure is also considered. A numerical example is given to illustrate the performance of the procedures. (kr)