Estimation of parameters of inverse Weibull distribution and application to multi-component stress-strength model. Issue 1 (2nd January 2022)

• Record Type: Journal Article
• Authors: Jana, Nabakumar; Bera, Samadrita
• Is Part Of: Journal of applied statistics. Volume 49:Issue 1(2022)
• Page Start: 169
• Page End: 194
• Publication Date: 2022-01-02
• Subjects: Maximum likelihood estimators -- equivariant estimator -- Bayes estimators -- ordered scale parameters -- UMVUE -- stress-strength reliability -- Monte Carlo simulations; 62F10 -- 62F30 -- 62N05; Statistics -- Periodicals; 519.5
• DOI: 10.1080/02664763.2020.1803815
• Languages: English
• ISSNs: 0266-4763

Abstract

Abstract : The problem of estimation of the parameters of two-parameter inverse Weibull distributions has been considered. We establish existence and uniqueness of the maximum likelihood estimators of the scale and shape parameters. We derive Bayes estimators of the parameters under the entropy loss function. Hierarchical Bayes estimator, equivariant estimator and a class of minimax estimators are derived when shape parameter is known. Ordered Bayes estimators using information about second population are also derived. We investigate the reliability of multi-component stress-strength model using classical and Bayesian approaches. Risk comparison of the classical and Bayes estimators is done using Monte Carlo simulations. Applications of the proposed estimators are shown using real data sets.