LAUSR

In this paper we develop two pivotal quantities to construct prediction intervals for future two-parameter exponential lifetimes based on a random number of generalized order statistics (gOs) under a general set-up including progressive type II censored order statistics (pOs) with general scheme. Moreover, the expected value as well as the mean square error for the upper limit of predictive confidence interval (PCI) are obtained. Furthermore, a maximum likelihood predictor (MLP) for future exponential lifetimes is derived when the sample size is deterministic or random. Finally, a simulation study and an application to real data illustrate and corroborate the theoretical results.