• 统计研究中心
当前位置: 首页 > 系列讲座 > 正文

加州大学戴维斯分校蒋继明教授: MIXED MODEL PREDICTION and SMALL AREA ESTIMATION

光华讲坛——社会名流与企业家论坛第 5838 期

(线上讲座)

主题MIXED MODEL PREDICTION and SMALL AREA ESTIMATION

主讲人加州大学戴维斯分校蒋继明教授

主持人统计学院 林华珍教授

时间2020年7月6日(周一)上午11:00-12:00

直播平台及会议ID腾讯会议,940 188 300

主办单位:统计研究中心、数据科学与商业智能联合实验室和统计学院 科研处

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。1995年从加州大学伯克利分校取得博士学位。他的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、大数据智能、统计遗传学/生物信息学、药代动力学和渐近理论。他先后出版了五本专著,包括线性和广义线性混合模型及其应用(Springer 2007),统计学大样本理论(Springer 2010),栅栏方法(World Scientific 2016),混合效应模型的渐进分析:理论、应用和开放性问题(Chapman&Hall / CRC,2017)和稳健混合模型分析(World Scientific 2019) )。他曾担任过包括AoS、JASA等统计学国际顶级期刊的编委。他是美国AAAS、ASA和IMS的Fellow, 也是ISI的Elected Member。他是1998年ASA的Outstanding Statistical Application Award的共同获奖者;他也是2015年第一个荣获NISS的Alumni Achievement Award的共同获奖者。他在JASA、AoS、JRSSB等统计学国际顶级期刊上发表了很多高质量的论文。详情请见其个人主页:https://statistics.ucdavis.edu/people/jiming-jiang


内容提要:

Mixed model prediction (MMP) has a fairly long history starting with Henderson's early work in animal breeding (Henderson 1948). The field has since flourished, thanks to its broad applications in various fields. The traditional fields of applications include genetics, agriculture, education, and surveys. This is a field where the frequentist and Bayesian approaches in statistics found common grounds. Nowadays, new and challenging problems have been emerging from such fields as business and health sciences, in addition to the traditional fields, to which methods of MMP are potentially applicable. This talk provides an overview of the basic elements of MMP as well as some of its modern developments. Of particular interest is application of MMP in small area estimation (SAE), a field that studies effective use of survey data for inference about small geographical areas and subpopulations. Of particular interest are recent developments in robust SAE and associated measures of uncertainty. Examples of application are discussed, including estimation of poverty rates for school-age children in U.S. and prediction of incubation period of Covid-19. Part of the work is jointly with Xiaohui Liu and Haiqiang Ma of Jiangxi University of Finance and Economics.

混合模型预测(MMP)有相当长的历史,始于Henderson在动物育种方面的早期工作(Henderson 1948)。由于在各个领域的广泛应用,该领域得到了蓬勃发展。传统的应用领域包括遗传学、农业、教育和调查。这是统计学中的频率学派和贝叶斯学派找到共同点的领域。如今,除了传统领域外,MMP方法在商业和健康科学等领域也出现了新的具有挑战性的问题。本讲座提供了MMP的基本要素以及一些现代发展的概述。MMPSAE中的应用特别受关注,该领域研究有效利用调查数据来推断小地理区域和亚人群的领域。特别值得关注的是稳健的SAE及其关于不确定性的度量。讨论了应用实例,包括估算美国学龄儿童贫困率和预测Covid-19潜伏期。部分工作由江西财经大学的刘小惠(Xiaohui Liu) 和马海强 (Haiqiang Ma)共同完成。


上一条:香港大学 李国栋教授:High-dimensional vector autoregressive time series modeling via tensor decomposition

下一条:加州大学伯克利分校丁鹏副教授:Berry-Esseen bounds for design-based causal inference with possibly diverging treatment levels and varying group sizes