题目：Paradoxes and resolutions for semiparametric fusion of individual and summary data. 

　　摘要：Suppose we have available  individual  data from  an internal study  and various types of summary statistics from relevant external studies.  External summary statistics have been used as constraints on the internal data distribution, which promised to  improve the statistical inference in the internal data; however, the  additional use of  external summary data may lead to paradoxical results:  efficiency loss  may occur  if the uncertainty of the summary statistics is not negligible  and  estimation bias can emerge     if  they are obtained in a different population   from the internal study. We investigate these    paradoxical results in a semiparametric framework. We establish the semiparametric efficiency bound for estimating  a  general functional of the internal data distribution, which is shown to be  no larger than that using only internal data.  We   propose  a data-fused efficient estimator that achieves this bound so that the efficiency paradox is resolved.  This   data-fused  estimator is further regularized with adaptive lasso penalty  so that the resultant estimator  can achieve the same asymptotic distribution as the oracle one that uses only unbiased summary statistics, which resolves the bias paradox. Simulations and   application to  a Helicobacter pylori infection dataset are used to illustrate the proposed methods.

　　报告人简介：苗旺现为北京大学概率统计系研究员， 2008-2017年在北京大学数学科学学院读本科和博士，2017-2018年在哈佛大学生物统计系做博士后研究，2018年入职北京大学光华管理学院，2020年调入数学科学学院。苗旺的研究兴趣包括因果推断，缺失数据分析，及其在生物统计，流行病学，经济学和人工智能研究中的应用，与合作者提出混杂分析的代理推断方法，发展非随机缺失数据的识别性和双稳健估计理论，以及数据融合的半参数理论。个人网页https://www.math.pku.edu.cn/teachers/mwfy/