Inference for the two-parameter exponential distribution with generalized order statistics.

Inferences about estimation and prediction of the two-parameter exponential distribution are based on generalized order statistics. Point and interval estimates are used for scale and location parameters. Unbiased point predictors and reconstructors are based upon pivotal quantities. The mean square error and Pitman's measure help assess the closeness of estimators and predictors. Point estimators of scale and location parameters and point predictors of future observations are computed in the application of durations until remission of 20 leukemia patients and in the application until failure of air-conditioning systems. [ABSTRACT FROM AUTHOR]