On Bayesian One-sample Prediction of the Generalized Pareto Distribution based on Generalized Order Statistics

Abstract

predictive functions for future observations from a generalized Pareto distribution based on generalized order statistics are obtained. Two cases are considered unknown one parameter and unknown two parameters. We also consider two cases fixed sample size and random sample size. The Bayesian predictive functions are specialized to ordinary order statistics, progressive type II censoring and upper record values. Examples are calculated for the lower and the upper bounds for the future observation based on ordinary order statistics, progressive type II censoring and upper record samples.. MSC 2000: 60G25; 15A09; 62J05 Keywordsprediction, generalized order statistics, generalized Pareto distribution, ordinary order statistics, predictive function, random sample size. k r k s r s s r r r s r s ) ( 1 ) ( s i r s a s i j r j i j r i           