Bayes Estimation of a Distribution Function Using Ranked Set Samples

摘要: Aranked set sample (RSS), if not balanced, is simply a of independent order statistics gener- ated from the same underlying distribution F. Kvam and Samaniego (1994) derived maximum likelihood estimates F for general RSS. In many applications, including some in environ- mental sciences, prior information about available to supplement data-based inference. such cases, Bayes estimators should be considered improved estimation. estimation (using squared error loss function) unknown function investigated with samples. Additionally, generalized estimator (GMLE) derived. An iterative scheme based on EM Algorithm used produce GMLE For case loss, simple solutions are uncommon, procedure find solution estimate using Gibbs sampler illustrated. The methods illustrated data Natural Environmental Research Council Great Britain (1975), representing water discharge floods Nidd River Yorkshire, England