Analysis of a semiparametric mixture model for competing risks
| 详细信息 | 推荐本文 | |
- 作者：Angelica Hernandez-Quintero (12) angelicahernandezquintero@gmail.com
Jean-Fran莽ois Dupuy (2) dupuy@math.univ-toulouse.fr
Gabriel Escarela (1) ge@xanum.uam.mx - 关键词：Censored failure time data – ; Competing risks – ; Large ; sample properties – ; Maximum likelihood estimation – ; Mixture model – ; Multinomial logistic – ; Proportional hazards model
- 刊名：Annals of the Institute of Statistical Mathematics
- 出版时间：April 2011
- 出版年：2011
- 期刊代码：1_00203157
- 类别：et
- 卷：63
- 期：2
- 页码：305-329
- 数据来源：sp
摘要
Semiparametric mixture regression models have recently been proposed to model competing risks data in survival analysis. In particular, Ng and McLachlan (Stat Med 22:1097–1111, 2003) and Escarela and Bowater (Commun Stat Theory Methods 37:277–293, 2008) have investigated the computational issues associated with the nonparametric maximum likelihood estimation method in a multinomial logistic/proportional hazards mixture model. In this work, we rigorously establish the existence, consistency, and asymptotic normality of the resulting nonparametric maximum likelihood estimators. We also propose consistent variance estimators for both the finite and infinite dimensional parameters in this model.