加拿大阿卡迪亚大学数学与统计系Jianan Peng教授

窗体底2017年12月22日，星期五下午15:30在大学城广州大学行政东楼后座310。应广州大学经济与统计学院和岭南统计科学研究中心的邀请，加拿大阿卡迪亚大学数学与统计系Jianan Peng教授为各位学子带来了精彩的学术讲座--《Interval estimates for low-dose risk assessment with nonquantal data》——暨羊城讲坛第二十六讲。此次讲座由经济与统计学院熊健教授主持，相关专业的师生也参加了此次讲座。在这个讲座中，Jianan Peng教授提出了低剂量风险评估的广义置信区间。 仿真研究表明，所提出的置信区间具有满意的覆盖概率。

摘要:

Risk assessment studies where human, animal or ecological data are used to set safe low dose levels of a toxic agent are challenging as  study information is limited to high dose levels of the agent. However,   interest often exists in the low-risk  responses associated with doses closer to zero and careful statistical studies are needed to make appropriate inference on the low-dose effects from the high-dose  data. Piegorsch et al. ( Journal of Biopharmaceutical Statistics Vol 15: 17-31) discussed several interval estimates for low-dose risk assessment. In this talk, we propose generalized confidence intervals for low-dose risk assessment. Simulation studies demonstrate that the proposed confidence intervals have satisfying coverage probabilities.