An overview of Bayesian prediction of future record statistics using upper record ranked set sampling scheme

Two sample prediction is considered for a one-parameter exponential distribution. In practical experiments using sampling methods based on different schemes is crucial. This paper addresses the problem of Bayesian prediction of record values from a future sequence, based on an upper record ranked set sampling scheme. First, under an upper record ranked set sample (RRSS) and different values of hyperparameters, point predictions have been studied with respect to both symmetric and asymmetric loss functions. These predictors are compared in the sense of their mean squared prediction errors. Next, we have derived two prediction intervals for future record values. Prediction intervals are compared in terms of coverage probability and expected length. Finally, a simulation study is performed to compare the performances of the predictors. The real data set is also analyzed for an illustration of the findings.