讲座题目：Permutation test under rotating sampling plan with clustered data

主讲人：Professor Jiahua Chen   英属哥伦比亚大学

讲座时间：2020年8月4日（周二）上午10：00—12：00

参与方式：腾讯会议会议 ID：194 559 343；会议直播网址: https://meeting.tencent.com/l/yDaJnJZA5UiC

主讲人简介：

  Professor Jiahua Chen is currently a Canada Research Chair, Tier I, in statistical inference and a Professor in the Department of Statistics,  University of British Columbia, Vancouver, Canada. Dr. Jiahua Chen received his Master's degree from the Institute of System's Science in Jan 1985, and Ph.\ D degree from the University of Wisconsin-Madison in July 1990 under the supervision of Professor Jeff Wu. 

  His research interests include Finite Mixture Models, Statistical Genetics, Empirical Likelihood, Survey Methodology, Design of Experiment, and others.

        He has published his research in most major statistical journals such as The Annals of Statistics, The Journal of American Statistical Association, The Journal of the Royal Statistical Society, Biometrika and Technometrics.

       He served as the president of the International Chinese Statistical Association in 2005, and the president of the Section of Survey Methodology, the Statistical Society of Canada, 2007.

       He was a member of the Grant Selection Committee of the Natural Science and Engineering Research Council of Canada.

       He is an elected fellow of the Institute of Mathematical Statistics and the American Statistical Association.

       He is the recipient of the CRM-SSC Prize in Statistics for his outstanding contributions to the statistical sciences. 

       This prestigious award, jointly sponsored by the Statistical Society of Canada(SSC) and the Centre de recherches mathématiques de Montréal (CRM), is given each year to a Canadian statistician in recognition of outstanding contributions to the discipline during the recipient's first 15 years after earning a doctorate.

        He received the Gold Medal, the top prize of the Statistical Society of Canada in 2014, the International Chinese Statistical Association distinguished achievement award in 2016.

讲座摘要：

   Consider a population made of clusters of sampling units,evolving temporally, spatially or according to other dynamics. We wish to monitor the evolution of its means, variances,medians or other population parameters.

   In a real world, clustered data are often collected via a rotating plan for administrative convenience and informativeness. Under rotating plans, not only the observations in the same clusters are correlated, but also observations on the same unit collected in different occasions {\blue are  correlated}.

   Ignoring the correlation structure in the clustered data can lead to invalid inference procedures. Accommodating such cluster structures in parametric models is difficult or can lead to unacceptable level of misspecification risk.

   In this paper, we explore the exchangeability in clustered data collected via rotating sampling plan to develop a permutation scheme for testing various hypotheses of interest.

   We also introduce a semi-parametric density ratio model to facilitate the multiple population structure in rotating sampling plans.

   The combination ensures the validity of the inference methods while {\blue extracting} maximum information carried by the rotating sampling plan. Simulation study indicates that the proposed tests firmly control the type I error when data are clustered (or not).

   The use of the density ratio model improves the power of the tests. We also investigate the influence of the choice of basis function in the density ratio modeland conclude that the efficiency gain is widely observed.