含有治愈部分的可加危险模型的估计

Estimation of Surviavl Cure and Additive Hazards Models

最近可加危险(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的.基于混合治愈模型,有界累计危险治愈模型和"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模型的识别和参数估计也将在本文给出讨论.

Most 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.