This study considers the problem of estimating unknown parameters of the Burr XII distribution under classical and Bayesian frameworks when samples are observed in the presence of progressively type-I hybrid censoring. Under classical approach, we employ EM and stochastic EM algorithm for obtaining the maximum likelihood estimators of model parameters. On the other hand, under Bayesian framework, we obtain Bayes estimators with respect to different symmetric and asymmetric loss functions under non-informative and informative priors. In this regard, we use Tierney–Kadane and importance sampling methods. Asymptotic normality theory and MCMC samples are employed to construct the confidence intervals and HPD credible intervals. To improve the estimation accuracy shrinkage pre-test estimation strategy is also suggested. The relative efficiency of these estimators with respect to both classical and Bayesian estimators are investigated numerically. Our simulation studies reveal that the shrinkage pre-test estimation strategy outperforms the estimation based on classical and Bayesian procedure. Finally, one real data set is analyzed to illustrate the methods of inference discussed here.
Bayesian estimation SEM algorithm Importance sampling Tierney–Kadane method Shrinkage pre-test estimation Progressively type-I hybrid censored data
This is a preview of subscription content, log in to check access.
The authors are thankful the reviewers for their valuable comments and suggestions which significantly improved the earlier version of this paper.
Abdel-Hamid HH (2009) Constant-partially accelerated life tests for burr type-XII distribution with progressive type-II censoring. Comput Stat Data Anal 53:2511–2523MathSciNetCrossRefzbMATHGoogle Scholar
Al-Hussaini EK, Jaheen ZF (1992) Bayesian estimation of the parameters, reliability and failure rate functions of the Burr type XII failure model. J Stat Comput Simul 41:31–40MathSciNetCrossRefzbMATHGoogle Scholar
Arabi Belaghi R, Arashi M, Tabatabaey SMM (2015) On the construction of preliminary test estimator based on record values for Burr 12 model. Commun Stat Theory Methods 44:1–23CrossRefzbMATHGoogle Scholar
Burr IW, Cislak PJ (1968) On a general system of distributions: 1. it’s curve-shaped characteristics; 2. the sample median. J Am Stat Assoc 63:627–635Google Scholar
Calabria R, Pulcini G (1996) Point estimation under asymmetric loss functions for left-truncated exponential samples. Commun Stat Theory Methods 25:585–600MathSciNetCrossRefzbMATHGoogle Scholar
Chen MH, Shao QM (1999) Monte Carlo estimation of Bayesian credible and HPD intervals. J Comput Gr Stat 8:69–92MathSciNetGoogle Scholar
Chien TL, Yen LH, Balakrishnan N (2011) Exact Bayesian variable sampling plans for the exponential distribution with progressive hybrid censoring. J Stat Comput Simul 81:873–882MathSciNetCrossRefzbMATHGoogle Scholar
Cramer E, Balakrishnan N (2013) On some exact distributional results based on type-I progressively hybrid censored data from exponential distributions. Stat Methodol 10:128–150MathSciNetCrossRefzbMATHGoogle Scholar
Dempster AP, Laird NM, Rubin DB (1977) Maximum likelihood from incomplete data via the em algorithm. J R Stat Soc Ser B 39(1):1–38Google Scholar
Kibria BMG (2004) Performance of the shrinkage preliminary tests ridge regression estimators based on the conflicting of W, LR and LM tests. J Stat Comput Simul 74:793–810MathSciNetCrossRefzbMATHGoogle Scholar
Kibria BMG, Saleh AKME (2004) Preliminary test ridge regression estimatorswith Student’s t errors and conflicting test-statistics. Metrika 59:105124CrossRefGoogle Scholar
Kibria BMG, Saleh AKME (2005) Comparison between Han-Bancroft and Brook method to determine the optimum significance level for pre-test estimator. J Prob Stat Sci 3:293–303Google Scholar