Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4949336 | Computational Statistics & Data Analysis | 2017 | 16 Pages |
Abstract
Additive-multiplicative hazard model is a natural extension of the proportional hazard model and the additive hazard model in survival analysis. It is classical for applying the martingale estimating functions to estimate the regression parameters. However, the generalized moment method is employed to estimate the coefficients via synthesizing the auxiliary subgroup survival information. The estimators are established to be consistent and asymptotically normal. Furthermore, the method is more efficient than the famous martingale approach. In particular, these asymptotic variance-covariances are identical as the number of subgroups is equal to one. The large sample property of the Breslow estimator for the baseline cumulative hazard function is also investigated. Some extensive simulation studies are conducted to evaluate the finite-sample performances of the proposed method. A real data study is analyzed to show its practical utility.
Related Topics
Physical Sciences and Engineering
Computer Science
Computational Theory and Mathematics
Authors
Wenpeng Shang, Xiao Wang,