【摘要】This talk introduces a novel method for selecting main effects and a set of reparametrized predictors called conditional main effects (CMEs), which capture the conditional effect of a factor at a fixed level of another factor. CMEs represent interpretable, domain-specific phenomena for a wide range of applications in engineering, social sciences and genomics. The key challenge is in incorporating the grouped structure of CMEs within the variable selection procedure itself. We propose a new method, cmenet, which employs two principles called CME coupling and CME reduction to effectively navigate the selection algorithm. Simulation studies demonstrate the improved performance of cmenet over generic variable selection methods, such as Lasso and SparseNet. Applied to a gene association study on fly wing shape, cmenet not only provides improved predictive performance over existing selection techniques, but also reveals important insight on gene activation behavior, which could guide further experiments. (Paper to appear in JASA T&M. Joint work with Simon Mak.)

【报告人介绍】吴建福院士(Jeff C.F. Wu)是佐治亚理工大学工业与系统工程学院可口可乐讲座教授、美国工程院院士，曾获COPSS Presidents’ Award、Shewhart Medal、COPSS R.A. Fisher Lecture Award、Deming Lecture Award等奖。吴院士1971年获得国立台湾大学数学学士学位，1976年获得加州大学伯克利分校统计学博士学位，曾在密西根大学、威斯康星大学、滑铁卢大学任教。