Time: 09:30-10:30am, June. 25th, 2019, Tuesday
Venue: RoomA1514, Science Building, North Zhongshan Road Campus
Spreker: Professor Yufeng Liu, University of North Carolina at Chapel Hill
Abstract: Gaussian graphical models are widely used to represent conditional dependence among random variables. In this talk, we consider high dimensional estimation of multiple graphical models arising from heterogeneous observations. An appealing feature of our methodology is to learn clustering structure while estimating graphical models. This is achieved via a high dimensional EM algorithm. A joint graphical lasso penalty is imposed to extract a common structure shared across all clusters. In theory, a non-asymptotic estimation error bound is derived for understanding the trade-off between statistical accuracy and computational complexity in the regularized Gaussian mixture models. Simulation studies and an application to a Glioblastoma cancer dataset further demonstrate the effectiveness of the proposed method.
Speaker’s Bio：Dr. Yufeng Liu is a Professor in statistics in Department of Statistics and Operations Research, Department of Genetics, and Department of Biostatistics at University of North Carolina at Chapel Hill. His research interests include statistical learning techniques for complex and high dimensional data, graphical models, and individualized decision rules. He is an elected fellow at American Statistical Association (ASA), Institute of Mathematical Statistics (IMS), and an elected member at International Statistical Institute (ISI).
