Statistical Inference for the Inverted Kumaraswamy Accelerated Model Under Type-I Generalized Hybrid Censoring with Applications

Abstract

This study investigates the methodologies for robust parameter estimation within the context of the parameters of the inverted Kumaraswamy model using data derived from step-stress partially accelerated life testing with Type-I generalized hybrid censoring. We formulate estimation procedures within both frequentist (maximum likelihood) and Bayesian frameworks, including the construction of asymptotic and credible intervals. Subsequently, we provide a formal derivation of the associated asymptotic and bootstrap confidence intervals. To address the analytical intractability of the Bayesian estimation, we employ Markov Chain Monte Carlo techniques. The proposed methods are illustrated through an illustrative example, an application to real-world precipitation data, and a simulation study.