报告题目：A general joint modeling framework for longitudinal and time-to-event data with flexible random-effects

报告人：黄希芬

时间：4月25日 14：50-15：40

地点：体育赛事直播平台 106会议室

报告摘要：Joint models are an increasingly popular way to characterise the relationship between one or more longitudinal responses and an event of interest. In general, statistical inference based on joint model requires accurate assumptions about the potential distribution of the random effects and their correlation structure. And even if we can accurately assume the potential distribution of the random effects, when the number of longitudinal measures, the complexity of the random effects structure, or both, grows, the model inference unavoidably involves multiple integrations which may be intractable and hence leads to severe computational challenges, especially in the presence of large longitudinal measures, hampering the implementation of many classic approaches. In this paper, we develop a general joint model framework without restricting the distribution form of the random effects, which solves the limitation that the existing methods are prone to model misspecification that leads to estimation bias. By combining the discrete approximation technique, an efficient minorization–maximization algorithm for the joint model is developed. Extensive simulation studies are conducted to demonstrate the scalability and accuracy of the proposed method. An application on the primary biliary cholangitis data is also presented to illustrate its practical utilities.

报告人简介：黄希芬，博士毕业于香港大学，现为云南师范大学数学体育赛事直播 教授，主要从事生物医学中统计模型和快速算法等方面的研究，主持国家自然科学基金项目3项，在Statistica Sinica, Statistical Methods in Medical Research, Biometrical Journal等统计期刊发表论文20余篇。2019年和2023年先后获得云南省“兴滇英才支持计划”青年人才和云南省社会科学奖等。