The cause-specific hazard function plays an important role in developing the regression models for competing risks survival data. Proportional hazards and additive hazards are the commonly used regression approaches in survival analysis. Mostly, in literature, the proportional hazards model was used for parametric regression modelling of survival data. In this article, we introduce a parametric additive hazards regression model for survival analysis with competing risks. For employing a parametric model we consider the modified Weibull distribution as a baseline model which is capable to model survival data with non-monotonic behaviour of hazard rate. The estimation process is carried out via maximum likelihood and Bayesian approaches. In addition to Bayesian methods, a class of non-informative types of prior is introduced with squared error (symmetric) and linear-exponential (asymmetric) loss functions. The relative performance of the different estimators is assessed using Monte Carlo simulation. Finally, using the proposed methodology, a real data analysis is performed.
Additive hazard Competing risks Modified Weibull distribution Bayes estimation Non-informative priors MCMC
Birincil Dil | İngilizce |
---|---|
Konular | İstatistik |
Bölüm | İstatistik |
Yazarlar | |
Erken Görünüm Tarihi | 13 Mayıs 2023 |
Yayımlanma Tarihi | 31 Ekim 2023 |
Yayımlandığı Sayı | Yıl 2023 Cilt: 52 Sayı: 5 |