Malaysian Journal of Mathematical Sciences, June 2026, Vol. 20, No. 2
Constant-Stress Partially Accelerated Life Tests of Kumaraswamy Distribution with Generalized Type-I Hybrid Censoring
Ali Mousa, M. A. M., Ramadan, H. A., Farghal, A. W. A
Corresponding Email: alwageh-ahmed@science.sohag.edu.eg
Received date: 1 August 2025
Accepted date: 16 November 2025
Abstract:
This paper discusses various methods for estimating the parameters of the Kumaraswamy (Kw) distribution and the acceleration factor in constant stress partially accelerated life test (CSPALTs) using Type-I generalized hybrid censored data. The constant stress partially accelerated life test CSPALTs using censored data that is Type-I generalized hybrid is explained in detail. In classical inference, maximum likelihood (ML) estimates are derived for both the model parameters of the Kumaraswamy (Kw) and the acceleration factor. The Fisher Information Matrix (FIM) is utilized to derive the approximate confidence intervals (ACIs) for all model parameters and the acceleration factor. The model parameters and the acceleration factor are estimated using bootstrap approach. Markov chain Monte Carlo (MCMC) method is suggested for Bayesian inference in order to optimize the related credible intervals (CRIs) and the Bayes estimators. In addition to, symmetric loss functions as well as asymmetric ones are considered. A collection of simulated data is analyzed to demonstrate the viability and appropriateness of the proposed methods. Furthermore, a Monte Carlo simulation is conducted to assess the effectiveness of the proposed methods.
Keywords: Kumaraswamy distribution; constant-stress partially accelerated life tests; generalized Type-I hybrid; ML estimation; bootstrap; MCMC.
