Order Restricted Inference for Adaptive Progressively Censored Competing Risks Data

TL;DR

This paper develops and compares order restricted and unrestricted inference methods for Weibull-based competing risks data under adaptive progressive Type-II censoring, demonstrating improved efficiency with order restrictions through simulations and a real example.

Contribution

It introduces Bayesian and likelihood-based inference techniques for order restricted Weibull competing risks data under adaptive censoring, highlighting efficiency gains over unrestricted methods.

Findings

Order restricted inference yields more efficient estimates.

Simulation confirms improved accuracy with order restrictions.

Method applied successfully to real data example.

Abstract

Under adaptive progressive Type-II censoring schemes, order restricted inference based on competing risks data is discussed in this article. The latent failure lifetimes for the competing causes are assumed to follow Weibull distributions, with an order restriction on the scale parameters of the distributions. The practical implication of this order restriction is that one of the risk factors is dominant, as often observed in competing risks scenarios. In this setting, likelihood estimation for the model parameters, along with bootstrap based techniques for constructing asymptotic confidence intervals are presented. Bayesian inferential methods for obtaining point estimates and credible intervals for the model parameters are also discussed. Through a detailed Monte Carlo simulation study, the performance of order restricted inferential methods are assessed. In addition, the results are also compared with the case when no order restriction is imposed on the estimation approach. The simulation study shows that order restricted inference is more efficient between the two, when this additional information is taken into consideration. A numerical example is provided for illustrative purpose.