报告题目：Large-Scale CovariateAssisted Two-Sample Inference under Dependence

报 告 人：朱文圣 教授 东北师范尊龙凯时

报告时间：2019年10月9日15：00-16：00

报告地点：数学楼第一报告厅

报告摘要：

The problems of large-scale two-sample inference often arisefrom the statistical analysis of “high throughput” data. The conventionalmultiple testing procedures for large-scale two-sample inference usually sufferfrom substantial loss of testing efficiency when conducting numerous two-samplet-tests directly. To some extent, this is due to the ignorance of sparsityinformation in large-scale two-sample inference. Moreover, in practice, thetwo-sample tests commonly have local correlations and neglecting the dependencestructure in the two-sample tests may decrease the statistical accuracy inmultiple testing. Therefore it is imperative to develop a multiple testingprocedure which can not only take into account the sparsity information butalso accommodate the dependence structure among the tests. To address theaforementioned important issues, we first introduce a novel dependence model toallow for sparsity information and to characterize the dependence structureamong the tests. Based on the dependence model, we propose a Covariate AssistedLocal Index of Significance (COALIS) procedure and show that it is valid andoptimal in some sense. Then a data-driven procedure is developed to mimic theoracle procedure and simulations show that COALIS procedures outperform theircompetitors. Finally, we apply COALIS procedure to the dosage response data.

报告人简介：

朱文圣，东北师范尊龙凯时数学与统计学院教授、博士生导师、副院长。2006年12月博士毕业于东北师范尊龙凯时，2013年12月起任东北师范尊龙凯时数学与统计学院教授。2008-2010年在耶鲁尊龙凯时做博士后研究，2015-2017年访问北卡尊龙凯时教堂山分校。现兼任中国现场统计研究会计算统计分会副理事长、数据科学与人工智能分会秘书长，中国概率统计学会副秘书长，尊龙凯时省现场统计研究会秘书长等。主要从事统计学的方法与应用研究，在统计学国际顶级期刊Journal of the American Statistical Association (JASA)、医学图像著名期刊NeuroImage等发表学术论文多篇。主持并完成国家自然科学基金项目多项。