讲座题目：Individualized Multi-directional Variable Selection

主讲人：Professor Annie Qu  加州大学欧文分校

讲座时间：2020年8月10日（周一）上午10：00—12：00

参与方式：会议 ID：292 419 730

会议直播: https://meeting.tencent.com/l/UfOqiINQycWe

主讲人简介：

Chancellor’s Professor, Department of Statistics, University of California Irvine Ph.D, Statistics, the Pennsylvania State University.

Qu’s research focus on solving fundamental issues regarding unstructured large-scale data, developing cutting-edge statistical methods and theory in machine learning and algorithms on text sentiment analysis, automatic tagging and summarization, recommender systems, tensor imaging data and network data analyses for complex heterogeneous data, and achieving the extraction of essential information from large volume high-dimensional data. Her research has impacts in many different fields such as biomedical studies, genomic research, public health research, and social and political sciences.

Before she joins the UC Irvine, Dr. Qu is Data Science Founder Professor of Statistics, and the Director of the Illinois Statistics Office at the University of Illinois at Urbana-Champaign. She was awarded as Brad and Karen Smith Professorial Scholar by the College of LAS at UIUC, a recipient of the NSF Career award in 2004-2009, and is a Fellow of the Institute of Mathematical Statistics and a Fellow of the American Statistical Association.

讲座摘要：

In this paper we propose a heterogeneous modeling framework which achieves individualwise feature selection and heterogeneous covariates’ effects subgrouping simultaneously.

In contrast to conventional model selection approaches, the new approach constructs a separation penalty with multi-directional shrinkages, which facilitates individualized modeling to distinguish strong signals from noisy ones and selects different relevant variables for different individuals. 

Meanwhile, the proposed model identifies subgroups among which individuals share similar covariates’ effects, and thus improves individualized estimation efficiency and feature selection accuracy. 

Moreover, the proposed model also incorporates within-individual correlation for longitudinal data to gain extra efficiency. We provide a general theoretical foundation under a double-divergence modeling framework where the number of individuals and the number of individual-wise measurements can both diverge, which enables inference on both an individual level and a population level.

In particular, we establish a strong oracle property for the individualized estimator to ensure its optimal large sample property under various conditions. 

An efficient ADMM algorithm is developed for computational scalability. Simulation studies and applications to post-trauma mental disorder analysis with genetic variation and an HIV longitudinal treatment study are illustrated to compare the new approach to existing methods. This is joint work with Xiwei Tang and Fei Xue.