We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
A New Two-Parameters Lindley-Frailty Model: Censored and Uncensored Schemes under Different Baseline Models: Applications, Assessments, Censored and Uncensored Validation Testing.
- Authors
Teghri, Samia; Goua, Hafida; Loubna, Hamami; Butt, Nadeem S.; Khedr, Abdelrahman M.; Yousof, Haitham M.; Ibrahim, Mohamed; Salem, Moustafa
- Abstract
Classical survival models assume homogeneity among the population of individuals who are susceptible to the event of interest. However, in many practical circumstances, there is a certain amount of unobserved heterogeneity that can be caused by a variety of sources, such as environmental or genetic factors. If the heterogeneity is ignored, many issues could arise, including an overestimation of the hazard rate and inaccurate estimates of the regression coefficients. Frailty models are usually used to model the heterogeneity among individuals. In this paper, we propose a novel univariate frailty model. The frailty variable is assumed to follow the Two Parameter Lindley distribution. The maximum likelihood method is used to estimate the model parameters. The baseline hazard functions are assumed to follow Weibull, Exponential, Gompertz, and Pareto distributions, and a simulation study is performed under this assumption. We examine the characteristics of the distribution and assess its performance compared to other distributions that are frequently applied in frailty modeling by using both Nikulin-Rao-Robson and Bagdonavicius-Nikulin goodness-of-fit tests to determine the adequacy of the model. We analyze a fresh medical dataset collected from an emergency hospital in Algeria to evaluate the effectiveness and applicability of the proposed model.
- Subjects
ALGERIA; FAILURE time data analysis; MAXIMUM likelihood statistics; PARETO distribution; GOODNESS-of-fit tests; HAZARD function (Statistics); CENSORSHIP
- Publication
Pakistan Journal of Statistics & Operation Research, 2024, Vol 20, Issue 1, p109
- ISSN
1816-2711
- Publication type
Article
- DOI
10.18187/pjsor.v20i1.4225