题    目：Bayesian optimal designs for linear regression with high predictive efficiency in the event of model uncertainty

主讲人：Po Yang 教授

单    位：University of Manitoba, Canada

时    间：2025年5月28日 10:00

地    点：学院二楼会议室

摘    要：It is well known that classical optimal designs depend on an assumed model which may not be a true model. To overcome this, Bayesian designs were introduced. For response surface experiments, the prediction of the response is an important task. Bayesian I-optimal designs minimize the average variance of prediction, thereby increasing the prediction efficiency. We introduce three new Bayesian optimality criteria for constructing optimal designs that have high prediction efficiency and less dependence on an assumed model. Optimal designs are obtained using the Bayesian I-criterion and the newly introduced criteria. The performance of the criteria is compared with that of existing optimality criteria using graphical methods and some efficiency measures.

简    介：杨钋在麦克马斯特大学获得博士学位, 曾任教于芝加哥的德保罗大学。现在是加拿大曼尼托巴大学统计系终身教授，副系主任，主管研究生工作。研究方向是实验设计与数据分析, 区组设计，响应优化和贝叶斯方法。曾在Statistica Sinica 和 Journal of Quality Technology等国际知名期刊发表学术论文，她的研究一直受到加拿大囯家自然科学和工程研究基金的资助。现任Taylor &Francis online 旗下的杂志Communications in Statistics编委。