报告题目：Communication-Efficient Distributed Subgroup Learning

报告人：朱文圣 教授 云南大学

报告时间：2025年9月10日 9:00-10:00

报告地点：腾讯会议ID：323-379-001

校内联系人：赵世舜 [email protected]

报告摘要：

Exploring the latent clustering structure of objects is a primary and fundamental task in many fields for massive data. For example, the aim of precision medicine is transforming traditional population-average treatment effects into individualized treatment effects, which usually need to identify subgroups from a heterogeneous population first. Massive data is frequently dispersed across multiple sites in the form of local generation, local collection, and local storage, which brings new challenges to subgroup analysis due to the issue of computing burden and communication costs. In this article, we study the subgroup analysis for distributed environment scenarios to identify subgroups of individuals from multiple different sites. To achieve efficient communication and privacy-protected grouping and estimation, we develop a distributed surrogate fusion penalized regression (DSFPR) approach, which consists of two stages. In the first stage, we construct a preliminary grouping structure through local subgroup analysis on each site. In the second stage, we propose the surrogate objective function based on the grouping structure obtained in the first stage, and perform global subgroup analysis. To address parallel problem-solving, we design a distributed alternating direction method of multiplier algorithm, which does not involve the transmission of personal information. We introduce the sub-oracle property for estimation in local subgroup analysis and establish theoretical properties for the final estimation under both correct and incorrect preliminary grouping structures. Finally, simulations and real data analysis validate the effectiveness of our approach.

报告人简介：

朱文圣，云南大学数学与统计学院教授、博士生导师。2006年博士毕业于东北师范大学，2008-2010年在耶鲁大学做博士后研究，2015-2017年访问北卡罗莱纳大学教堂山分校。研究方向为生物统计学及统计机器学习，在JASA, Biometrika, Statistica Sinica，Science China-Mathematics，The Lancet等杂志发表学术论文多篇，主持多项国家级科研项目。中国现场统计研究会贝叶斯统计分会理事长、中国统计教育学会副会长，全国工业统计学教学研究会副会长。国家级一流本科课程负责人。