一、报告时间

2025年05月21日15:00

二、报告形式

腾讯会议号：427545253

三、报告人

黄振生教授   南京理工大学

四、报告主题

Scaled Partial Envelope Model in Multivariate Linear Regression

五、报告摘要

Inference which is based on the partial envelope model is variational or nonequivariant under rescaling of the responses, and it is inclined to restrict utilization to responses that are measured in the identical or analogous units. The efficiency acquisitions promised by partial envelopes frequently cannot be accomplished when the responses are measured in diverse scales. Based on the research results of Cook and Su (2013), we extend the partial envelope model to the scaled partial envelope model which can overcome the disadvantage above and enlarge the scope of partial envelopes. It maintains the potential of the partial envelope model to add up efficiency and is invariable to scale changes. Further, we demonstrate the maximum likelihood estimators and their properties. At last, simulation studies and real data example demonstrate that the advantage of scaled partial envelope estimators, including a comparision with the standard model estimator, the partial envelope estimator and the scaled envelope estimator.

六、报告人简介

黄振生，理学博士，南京理工大学数学与统计学院教授、博士生导师，研究方向包括非参数与半参数统计推断、大数据中的统计方法及应用等。主持多项国家及省部级项目，发表学术论文70余篇。曾获得省部级优秀博士论文、省级中青年学术带头人、澳大利亚联邦政府“奋进”奖。