|
|
 |
| |
|
|
|
Statistical Inference and Applications of the Inverse Power Half-Logistic Distribution in Reliability Analysis |
|
|
|
PP: 819-836 |
|
|
doi:10.18576/amis/200319
|
|
|
|
Author(s) |
|
|
|
Abdallah G. Bani-Saleh,
Dina A. Ramadan,
B. S. El-Desouky,
Mahmoud M. El-Awady,
|
|
|
|
Abstract |
|
|
| This paper proposes a new two-parameter
life time model, referred to as the Inverse Power Half-Logistic (IPHL)distribution, which is constructed by applying an inverse power transformation to the power half-logistic distribution. The IPHL model demonstrates high flexibility and is capable of modeling various hazard rate shapes, including increasing, decreasing, unimodal, and bathtub-shaped patterns, making it suitable for reliability and survival data analysis. Key statistical and reliability properties of the proposed distribution are derived, such as the density and distribution functions, survival and hazard functions, moments, quantile function, mean residual life, Re ́nyi entropy, and stress–strength reliability measure. Parameter estimation is carried out using several methods, including maximum likelihood estimation (MLE), maximum product spacing (MPS), and least squares estimation (LSE). In addition, Bayesian estimation (BE) is performed using Markov Chain Monte Carlo (MCMC) techniques. Along with MPS and LSE methods. Interval estimation is addressed through asymptotic, bootstrap, and Bayesian credible intervals. A Monte Carlo simulation study is conducted to assess and compare the performance of the proposed estimators. Furthermore, the applicability of the IPHL distribution is illustrated using two real lifetime datasets, where its performance is compared with several competing inverse models. The results indicate that the IPHL distribution provides improved goodness-of-fit and greater modeling flexibility, confirming its effectiveness for practical reliability applications. |
|
|
|
|
|
|
|
