ECE Seminar: Yanting Ma

Yanting Ma Ph.D. candidate North Carolina State University Faculty Host: Vivek Goyal Light refreshments will be available at 9:45 am outside of PHO 339 Title: Solving Large-Scale Inverse Problems via Approximate Message Passing and Optimization Abstract: The problem of reconstructing a signal or an object from the acquired measurements is ubiquitous in various computational sensing problems. In some cases, a linear formulation is sufficient to model the sensing process, whereas in other cases, nonlinear formulation is needed. In the first part of the talk, I will introduce the application of a high-dimensional inference algorithm, approximate message passing (AMP), to several linear inverse problems including hyperspectral imaging, multiprocessor computing, and universal compressed sensing, which is then followed by rigorous performance analysis of AMP in a specific setting. In the second part of the talk, I will briefly discuss some background of an ongoing work about multiple scattering inversion based on optimization with a nonlinear forward model, hence a nonlinear inverse problem. Bio: Yanting Ma is a Ph.D. candidate in electrical engineering (with minor in applied mathematics) at North Carolina State University. Her research interests are mainly in low complexity algorithm design and analysis for large-scale linear and nonlinear inverse problems using approximate message passing and optimization techniques. Additionally, she is generally interested in probability theory, stochastic processes, convex analysis, and partial differential equations. She received the B.S. degree in communication engineering from Wuhan University, China in 2012 and was summer intern at Mitsubishi Electric Research Laboratories (MERL) in 2016 and 2017.