Analysing load-sharing system under progressive censoring using statistical and machine learning approach

This study focuses on the reliability analysis of a multi-component load-sharing system under a progressive censoring scheme. In such a system, where a component fails, its load is redistributed among the surviving components, increasing their failure rates. Such systems are known as load-sharing systems. The lifetimes of the components follow the Weibull distribution, we estimate system parameters and reliability measures using both classical and Bayesian frameworks. In the classical approach, the maximum likelihood estimation method is employed, with asymptotic confidence intervals derived for system parameters. For the Bayesian framework, we use informative priors to obtain Bayesian estimates and construct the highest posterior density intervals. Monte Carlo simulation study is presented to compare the performance of these estimation methods. Furthermore, real data is used to validate the proposed methodologies, demonstrating the effectiveness of the approaches. Furthermore, we introduce machine learning techniques to analyze the reliability of the model. We have employed three different machine learning methods, such as support vector regressor, random forest regressor, and gradient boosting regressor, to estimate reliability. A comparative analysis has been discussed of all three methods by using performance metrics. Finally, the influence of model parameters on reliability estimation has been explored.