主 题：A Pairwise-Likelihood Augmented Estimator for Cox Model Under Left-Truncation
主讲人：Prof. Yi Li
主持人：林华珍教授
时 间：2015年7月14日下午2点-3点
地 点：通博楼B212学术会议室
主办单位：统计研究中心 统计学院 科研处
主讲人简介：
Yi Li 密歇根大学教授，Kidney Epidemiology and Cost Center主任，NSF Statistics and Probability Program的评审专家，是包括统计学顶级期刊Journal of the American Statistical Association、Biometrics、Scandinavian Journal of Statistics在内的五个期刊副主编。 李教授从事统计以及生物统计研究，主持美国国家癌症研究院等基金10多项，基金总计超过2400多万美元，发表论文100余篇，其中在国际顶级杂志Journal of the American Statistical Association、Biometrics、Biometrika、Journal of the Royal Statistical Society:Series B等期刊发表论文20多篇。 为国际生物统计学会，美国统计学会（ASA Fellow，约千分之三的美国统计学会会员获此荣誉），国际数理统计学会以及泛华统计协会会员。
内容提要：
Survival data collected from prevalent cohorts are subject to left-truncation. Conventional conditional approaches can be inefficient by ignoring the information in the marginal likelihood of the truncation time. On the other hand, the stationarity assumption under length-biased sampling methods will lead to biased estimation when it is violated. In this paper, we propose a semiparametric estimation method by augmenting the Cox partial likelihood with a pairwise likelihood. We eliminate the unspecified truncation distribution in the marginal likelihood, yet retain the information about regression coefficients and the baseline hazard. Self-consistency of the estimator guarantees a fast algorithm to solve for the regression coefficients and the baseline hazard simultaneously. The proposed estimator is shown to be consistent and asymptotically normal with a consistent sandwich-type variance estimator. Simulations show a substantial efficiency gain in both the regression coefficients and the cumulative hazard over Cox estimators, and that the gain is comparable to LBS methods when the uniform truncation assumption holds. A data analysis illustrates the application of the methods.
