An Empirical Likelihood With Estimating Equation Approach for Modeling Heavy Censored Accelerated Life-Test Data

This article uses empirical likelihood with estimating equations to model and analyze heavy censored accelerated life testing data. This approach flexibly and rigorously incorporates distribution assumptions and regression structures into estimating equations in a nonparametric estimation framework. Real-life examples of using available data to explore the regression functional relationship and distribution assumption are provided. Derivation of asymptotic properties of the proposed method provides an opportunity to compare its estimation quality to commonly used parametric MLE methods in the situation of misspecified regression models. These real-life examples and asymptotic studies show a significant potential of the proposed methodology.