报告题目：Poisson Nonlinear Mixed Models for Dose-Response Curve Data

报告摘要：There is vast literature on Cytogenetic Dose-Response Curve studies where the number of genetic damages has been repeatedly measured in human lymphocytes exposed to radiation over subjects with varying dose levels and dose rates. The goal is to link genetic damages with both dose levels and dose rates. A linear-quadratic mean function of dose has been traditionally fitted with Poisson models for such data; however, this linear quadratic mean function has been found inadequate except for very low doses. In addition, this approach does not account for overdispersion in Cytogenetic Dose-Response data appropriately. In this paper, we consider Poisson nonlinear mixed models for Dose Response data. These models allow us to characterize the effects of both dose level and dose rate on genetic damages at the same time while accounting for overdispersion in the data.

报告人简介：马人俊教授，1999年毕业于加拿大英属哥伦比亚大学，获统计学系博士学位，1998年-2000年在渥太华大学作为生物统计学家和博士后工作，2000年开始在新不伦瑞克大学工作。曾经是加拿大Junior Research Chair, 2009年获终身教授，是北大西洋数学科学学会资助的有世界影响的多学科合作研究组“复杂相关数据的统计模型及其应用”的学科带头人。主要研究领域包括：随机效应广义线性模型，纵向数据与空间数据分析，环境与生物统计，零点膨胀相依数据分析。他最近与中国的合作者一起开发了新方法来解决地震、股票和保险研究中的问题。

报告时间：2025年6月20日（周五）15:00—17:00

报告地点：同析4号楼308室