摘要
最近以来,可加危险(AH)模型被广泛地应用于生存分析数据,模型的协变量可以假设为时间独立或时间相关.基于混合、非混合治愈模型和“不正确”的比例危险模型,本文将上述的可乘危险模型延伸到可加的危险模型,这里的模型可以允许含治愈部分的生存数据的存在.“不正确”的AH模型的模型识别和参数估计也在文中进行了讨论.
Recently, additive hazards (AH) models are extensively applied to analyze survival data, in which covariates are assumed to be time - independent or time - dependent. In this paper, we extend multiplicative hazards model which are based on mixture cure, no - mixture cure models and “improper” proportional hazards models to AH model which may accommodate survival data with long - term survivors. Identifiability of “improper” AH and its estimation are investigated in this paper.
出处
《重庆文理学院学报(自然科学版)》
2006年第4期47-51,共5页
Journal of Chongqing University of Arts and Sciences
关键词
可乘危险
可加危险
治愈模型
模型识别
部分似然
Multiplicative heazrds
additive hazards
cure model
model identifiability
partial likelihood