报告题目：Regression analysis of group-tested multivariate current status data

报告时间：2026年7月21日上午11:00

报告地点：南湖校区综合三报告厅

主办单位：激情视频

报告人：李树威

报告人简介：李树威, 统计学博士, 现任广州大学经济与统计学院教授、博士生导师。博士毕业于吉林大学统计系。主要研究方向为生物统计、大数据处理及机器学习，在Biometrika、Biometrics、Statistics in Medicine、Statistica Sinica、JCGS等期刊上发表论文40多篇。主持国家自然科学基金面上项目、国家自然科学基金青年项目、广东省自然科学基金面上项目、广州市科技局项目等

摘要：In large-scale disease surveillance, group testing, where individual specimens are pooled to test for the presence of a disease, is often used to save the screening time and cost. When group testing with a primary focus on the disease onset time is conducted to monitor multiple diseases simultaneously, group-tested multivariate current status data arise. For analyzing this emerging type of survival data, this work develops an efficient regression methodology with frailty proportional hazards models. By approximating each conditional cumulative baseline hazard function with monotone splines, we propose a sieve maximum likelihood approach and develop a stable EM algorithm to identify the sieve estimators. The asymptotic properties of the resultant estimators are established by utilizing the empirical process and sieve estimation theories in a novel way. The numerical results in the simulation studies demonstrate that the proposed method performs well in finite sample. An application to a sexually transmitted infection data set is provided for an illustration.