A New Log-Exponential-Exponential Accelerated Failure Time Model for Censored and Uncensored Data with Applications and Rigorous NRR Validation

Authors

  • Mohamed Ibrahim Department of Quantitative Methods, School of Business, King Faisal University, Al Ahsa 31982, Saudi Arabia
  • Hafida Goual Laboratory of probabilities and statistics LaPS, Badji Mokhtar-Annaba University, 12, P.O. Box, 23000 Annaba Algeria
  • Abdullah H. Al-Nefaie Department of Quantitative Methods, School of Business, King Faisal University, Al Ahsa 31982, Saudi Arabia
  • Sami A. Morsi Applied College, Abqaiq Branch, King Faisal University, P.O. Box 4000, Al-Ahsa 31982, Saudi Arabia
  • Haitham Yousof Department of Statistics, Mathematics and Insurance, Faculty of Commerce, Benha University, Egypt
  • Rehab Shehata Mahmoud Department of Statistics, Mathematics and Insurance, Faculty of Commerce, Benha University, Egypt

DOI:

https://doi.org/10.19139/soic-2310-5070-4838

Keywords:

Accelerated Failure Time; Censored Data; Exponential Distribution; Maximum Likelihood Estimators; Nikulin-Rao-Robson; Actuarial Risk Measures.

Abstract

The accelerated failure time (AFT) model is a cornerstone of survival analysis, yet traditional baselinedistributions often lack the flexibility required to model complex, non-monotonic hazard rates. This paper proposes anovel AFT model based on the log-exponential-exponential (LEE) distribution, offering enhanced flexibility for reliability,medical, and actuarial data analysis. We derive the fundamental mathematical properties of the LEE-AFT model and developthe maximum likelihood estimators (MLEs) for both uncensored and right-censored data schemes. To ensure rigorousmodel validation, we adapt the modified Nikulin-Rao-Robson (NRR) chi-square statistic to assess the goodness-of-fit whenparameters are estimated from complete and censored samples. Extensive Monte Carlo simulations are conducted to evaluatethe consistency, bias, and efficiency of the MLEs, as well as to verify that the empirical rejection rates of the NRR test closelyalign with nominal theoretical levels. The practical applicability of the proposed framework is demonstrated using four realworld datasets: electric insulating fluid, body fat, motor insulation, and emergency care data. Furthermore, we present acomprehensive actuarial case study using Norwegian fire insurance claims to evaluate risk measures such as Value-at-Risk(VaR) and Tail Value-at-Risk (TVaR). Fitted under log-linear, power-rule, and Arrhenius covariate assumptions, the AFTLEE model consistently provides superior fits compared to classical baselines, confirming its value as a robust tool forcomplex survival analysis, reliability engineering, and financial risk management.

Downloads

Published

2026-11-16

How to Cite

Ibrahim, M., Hafida Goual, Abdullah H. Al-Nefaie, Sami A. Morsi, Yousof, H., & Rehab Shehata Mahmoud. (2026). A New Log-Exponential-Exponential Accelerated Failure Time Model for Censored and Uncensored Data with Applications and Rigorous NRR Validation. Statistics, Optimization & Information Computing. https://doi.org/10.19139/soic-2310-5070-4838

Issue

Section

Research Articles

Categories

Most read articles by the same author(s)

1 2 > >>