学术活动
首页  -  科学研究  -  学术活动  -  正文
【学术报告】2021年7月5日姚远教授举办学术讲座

作者: 时间:2021-07-02 点击量:

报告人:姚远(香港科技大学

邀请人:刘海霞

报告时间:2021年7月5日(星期一)10:00-12:00

报告地点:腾讯会议:711 873 080

报告题目:Controlling the False Discovery Rate in Structural Sparsity: Split Knockoffs

报告摘要:Controlling the False Discovery Rate in variable selection is critical for reproducible discoveries, which receives extensive studies in sparse linear models. However, in many scenarios such as medical image analysis, partial order ranking, and trend filtering etc., sparsity constraint is not directly imposed on parameters, but on alinear transformation of the parameters to be estimated. In this talk, we present a new data adaptive variable selection framework with FDR control in this structural sparsity setting, the Split Knockoff. This proposed method adopts the variable splitting scheme from optimisation, which relaxes the linear subspace constraint on the parameters to its neighbourhood. Such a scheme will yield orthogonal design matrices and enjoy new statistical benefits in terms of FDR control and power. In particular, two variants of Split Knockoffs will be introduced to overcome the challenge of broken symmetry in exchangeability and achieve provable FDR control.

报告人简介:Prof. Yuan Yao is currentlyAssociate Professor of Mathematics and Chemical & Biological Engineering atHong Kong University of Science and Technology. He received Ph.D. in Mathematics from UC Berkeley and has been working on mathematics of machine learning and data science.




地址:中国·湖北省·武汉市珞喻路1037号 华中科技大学逸夫科技楼南楼
邮政编码:430074     电话\传真:86-027-87543231 E-mail:mathhust@mail.hust.edu.cn
Copyright 2021 ◎ 华中科技大学数学与统计学院