Prediction of times to failure of censored units under generalized progressive hybrid censoring scheme

• Published in: Computational statistics Vol. 37; no. 4; pp. 2049 - 2086
• Main Authors: Ahmadi, J., Khatib Astaneh, B., Rezaie, M., Ameli, S.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01-09-2022; Springer Nature B.V
• Subjects: Algorithms; Economic Theory/Quantitative Economics/Mathematical Methods; Experiments; Failure; Mathematics and Statistics; Original Paper; Probability and Statistics in Computer Science; Probability Theory and Stochastic Processes; Random variables; Statistics; Survival analysis; Weibull distribution; Best unbiased predictor; Mean squared prediction error; Conditional median predictor; Monte-Carlo simulation; Maximum likelihood predictor; Prediction interval
• ISSN: 0943-4062; 1613-9658
• DOI: 10.1007/s00180-021-01191-3

In this paper, the problem of predicting times to failure of units censored in multiple stages of generalized progressively hybrid censoring from exponential and Weibull distributions is discussed. Different classical point predictors, namely, the best unbiased, the maximum likelihood and the conditional median predictors are all derived. Moreover, the problem of interval prediction is investigated. Numerical example as well as two real data sets are used to illustrate the proposed prediction methods. Using a Monte-Carlo simulation algorithm, the performance of the point predictors is investigated in terms of the bias and mean squared prediction error criteria. Also, the width and the coverage rate of the obtained prediction intervals are studied by simulations.