Bayesian Reliability Analysis of a Sine Model Under Hybrid Censoring

This work introduces a three-parameter sine Kumaraswamy Shanker distribution (SKSD), a novel extension of the Shanker distribution that seeks to further enhance the flexibility of statistical models in explaining many real-life patterns of data. Through extending the Shanker distribution, the SKSD overcomes the deficiencies of the Shanker distribution, particularly in the modeling of advanced behaviors hazard rate functions. The suggested distribution can represent numerous real-world phenomena due to its different increasing and decreasing hazard rate shapes. Crucial statistical properties such as the quantile function, expansion of the density function, ordinary and incomplete moments, and stress-strength reliability are considered. The model parameters are estimated using both classical and Bayesian methods under adaptive Type-II hybrid progressive censoring with random unit removal. Additionally, the Fisher information matrix is used to derive asymptotic confidence intervals, while Markov Chain Monte Carlo methods are applied to estimate Bayesian credible intervals under symmetric and asymmetric loss functions. It is impossible to compare theoretically estimating processes. In order to investigate the behavior of various estimators and enable comparison using some criteria, extensive simulation research is performed. According to empirical results, relative absolute biases and mean squared errors tend to decrease with increasing sample size. The Bayesian estimate approach typically yields more accurate and stable findings than the maximum likelihood estimate. The SKSD’s adaptable features make it competitive for fitting jute fiber data when compared to the other alternatives.