报告题目：Assessing Uncertainty in Small Area Estimation under A Misspecified Model

主讲人：蒋继明教授（加州大学戴维斯分校）

时间：2024年7月12日（周五）15:30 p.m.

地点：北院卓远楼305会议室

主办单位：统计与数学学院

摘要：Observed best prediction (OBP) is a method of small area estimation that is known to be more robust against model misspecification than the traditional empirical best linear unbiased prediction method. However, assessing uncertainty in OBP has been a difficult task due to the potential model misspecification. This is because the assumed model cannot be used in deriving an uncertainty measure for OBP (otherwise, it would defeat the whole purpose of OBP). This talk provides an overview of a series of methods that have been developed for estimating the mean squared prediction error of OBP. Most of the developments focus on area-level models, where the robustness of OBP is well established. The models under consideration include the Fay-Herriot model and Poisson/gamma model for small area counts. Examples of empirical studies and real-data applications are discussed. This work is joint with Xiaohui Liu, Haiqiang Ma of Jiangxi University of Finance and Economics, China and Thuan Nguyen of Oregon Health and Science University, USA.

主讲人简介：

蒋继明，现为加州大学戴维斯分校的统计学教授, 统计系系主任。其研究兴趣包括混合效应模型、模型选择、小区域估计、纵向数据分析、精准医疗、大数据智能、隐私保护、统计遗传学/生物信息学、年龄标准化癌症率以及渐近理论；发表研究论文超过100篇，其中多篇刊在AoS、JASA、JRSSB和Biometrika等顶级统计与数据科学期刊上；先后出版了五本专著，包括《Generalized Linear Mixed Models and Their Applications》、《Large Sample Techniques for Statistics》、《The Fence Methods》等。蒋继明教授是AoS和JASA等多个统计学国际期刊的编委，是美国科学促进会（AAAS）、美国统计协会（ASA）、国际数理统计学会（IMS）的Fellow，也是国际统计学会（ISI）的Elected Member。