题 目：Conditional Quasi-likelihood Inference for Mean Residual Life Regression with Clustered Failure Time Data

报告人： Liming Xiang 副教授（南洋理工大学）

摘要:  In the analysis of clustered failure time data, Cox frailty models have been extensively studied by incorporating frailty with a prespecified distribution to address the potential correlation of data within clusters. In this work, we propose a frailty proportional mean residual life regression model to analyze such correlated data. A novel conditional quasi-likelihood inference procedure is developed, which utilizes the stochastic process and the inverse probability of censoring weighting to form estimating equations for regression parameters. Our proposal employs conditional inference based on a penalized quasi-likelihood to address within-cluster correlation without specifying the frailty distribution. By adopting the Buckley-James estimator in the inverse probability of censoring weighting, the method further allows for dependent censoring. We establish the asymptotic properties of the proposed estimator and evaluate its finite sample performance via simulation studies. An application to the data from a multi-institutional breast cancer study is presented as an illustration.

报告人简介： Liming Xiang is an Associate Professor of Statistics at the School of Physical and Mathematical Sciences, Nanyang Technological University, Singapore. Her general research interests are in the development of statistical methods for survival data and longitudinal/clustered data with applications to scientific problems in the areas of medicine, epidemiology and public health. Her current research particularly focuses on semiparametric methods for analyses of clustered survival data, semi-competing risks data, interval-censored data and high-dimensional survival data. Dr. Xiang serves as an associate editor for Statistics in Medicine and Computational Statistics & Data Analysis.

报告时间：2024年5月22日上午10：00-11：00

报告地点：腾讯会议：685-548-055

联系人：胡涛