报告题目：Censored quantile regression based on multiply robust propensity scores

主 讲 人：秦国友教授

报告时间：2020年12月9日10:00-11:00

报告地点：腾讯会议 ID：535 285 490

点击链接入会：https://meeting.tencent.com/s/LQ1TJwvrBwe4

报告摘要：

Censored quantile regression has attracted extensive research interest in recent years. Estimation of censored quantile coefficients based on information subsets of samples requires estimation of propensity scores, which has been studied by using either parametric or nonparametric methods. The parametric methods have the risk of incorrectly specifying the propensity score model while the nonparametric approaches suffer the issue of "the curse of dimensionality". In this paper, we propose a new method for estimation of the propensity score which is robust against misspecification of parametric models. The proposed estimator is consistent if any one of those multiply models is correctly specified. We also construct an estimator for the censored quantile coefficient based on multiply robust propensity scores. Large sample properties of the proposed estimator, such as the consistency and the asymptotic normality, are thoroughly investigated. We also derived the consistent of the propensity score estimator. Extensive simulation studies are conducted to investigate the performance of the proposed estimator. As an application, we analyzed a data-set from a HIV study.

主讲人介绍：

秦国友，教授，博士生导师。复旦大学公共卫生学院生物统计学教研室主任。主要从事生物统计学方法学和应用研究，包括针对复杂数据、复杂统计模型的统计方法研究，以及生物统计学方法在医学和公共卫生领域的应用。在医学顶级期刊British Medical Journal（BMJ）和生物统计权威期刊Biometric, Biostatistics, Statistics in Medicine等学术期刊上发表80余篇研究论文。在纵向数据方面相关研究工作获得教育部高等学校科学研究优秀成果奖二等奖。担任中华预防医学会生物统计分会第一届青年委员会主任委员以及<<中国卫生统计>>编委。

欢迎各位老师同学积极参加！