سیدمحمدتقی کامل میرمصطفائی

In this paper, the Pitman closeness of predictors for a future order statistic coming from a two parameter exponential distribution is discussed. The optimum equivariant predictor using the Pitman closeness criterion is derived and compared with the best invariant predictor and the best unbiased predictor. The predictors are obtained on the basis of observed record values. Exact expressions for the required Pitman closeness probabilities are derived when both parameters are considered to be unknown. Similar results are obtained for the predictors of the mean of a future sample from exponential distribution. Several tables which contain numerical computations, are provided in order to compare the predictors in the sense of Pitman’s measure of closeness.