Structured Sparse Learning and Its Applications

Structured Sparse Learning and Its Applications

时间： 地点：九龙湖校区计算机学院四楼会议室

报告简介：

Structured sparse learning takes advantages of the (pre-defined) structure of the variables/features. The usage of such structure leads to improved prediction performance and better interpretability. In this talk, we are going to discuss (1) several structured sparse learning approaches including group Lasso, fused Lasso, tree structured group Lasso, etc.; (2) the efficient optimization of the non-smooth convex problem; and (3) the application of structured sparse learning in variable selection and parallel MR reconstruction. The structured sparse learning approaches to be discussed have been implemented in the SLEP package, which has been downloaded over 50,000 times and cited over 150 times.

报告人简介：

Jun received his bachelor's degree in computer science from Nantong Institute of Technology (now Nantong University) in 2002, and Ph.D. in computer science from Nanjing University of Aeronautics and Astronautics in 2007. Jun was a faculty member of Nanjing University of Aeronautics and Astronautics (2007.11-2010.5), a postdoc at Arizona State University (2008.2-2011.2) developing the SLEP package, and a Research Scientist at Siemens Corporate Research (2011.2-2012.12) developing parallel MR reconstruction approaches. Since 2013.1, Jun has worked at SAS Institute Inc. with focus on variable selection. Jun has published over 50 conference/journal papers which have been cited over 1000 times. Jun maintained a homepage at https://sites.google.com/site/junliupage/.

计算机网络和信息集成教育部重点实验室（东南大学）