Bayesian Estimation Strategy for a Newly Developed 2‐Component Mixture Model of Exponential Distributions Under Random Censoring Scheme

• Published in: Modelling and Simulation in Engineering Vol. 2025; no. 1
• Main Authors: Ullah, Zain, Ali, Amjad, Khan, Akbar Ali, Ali, Mehboob, Hussain, Ishtiaq, Shah, Rasool
• Format: Journal Article
• Language: English
• Published: New York John Wiley & Sons, Inc 01.01.2025; Wiley
• Subjects: Analysis; Bayesian analysis; Censorship; Experiments; Parameter estimation
• ISSN: 1687-5591; 1687-5605
• DOI: 10.1155/mse/3092633

Mixture models serve a crucial function in life testing experiments, particularly when dealing with a heterogeneous population. In our current research study, we have developed a 2‐component mixture model of the exponential distributions (2‐CMMEDs) for clinical experiments that involve random censoring. We have obtained closed‐form expressions for the Bayes estimators (BEs) and Bayes posterior risks (BPRs) for the parameters of 2‐CMMEDs under both informative (Gamma) and noninformative (Jeffreys’) priors and employing various loss functions. We investigate the performance of these BEs across different sample sizes and parametric values under different loss functions. Theoretical findings are further validated through simulation studies, as well as real data analysis. The numerical findings depict that the Gamma prior performs better, while the DeGroot loss function yields efficient results for estimating the parameters of a 2‐CMMEDs.