Loading Events
  • This event has passed.

Sparsity Constrained Estimation Using Spike and Slab Priors

Friday, August 12, 2016 @ 11:00 AM - 12:00 PM

Event Category:

August 12, 2016 at 11:00 a.m. Prof. Vishal Monga, Associate Professor at Pennsylvania State University, will be presenting “Sparsity Constrained Estimation Using Spike and Slab Priors”.

Speaker: Prof. Vishal Monga
Associate Professor, Pennsylvania State University, University Park

Day & Time: Friday August 12th, 2016
11:00 a.m. – 12:00 p.m.

Location: Room BA 7129 (tentatively)
Bahen Centre for Information Technology
40 St George St, Toronto, ON M5S 2E4

Contact: Eman Hammad

Abstract: We address sparse signal, i.e. image recovery in a Bayesian estimation framework where sparsity is enforced on reconstruction coefficients via probabilistic priors. In particular, we focus on the popular spike and slab prior which is considered the gold standard in the statistics literature. The optimization problem resulting from this model has broad applicability in recovery, regression and classification problems and is known to be a hard non-convex problem whose existing solutions involve simplifying assumptions and/or relaxations. We propose an approach called Iterative Convex Refinement (ICR) that aims to solve the aforementioned optimization problem directly allowing for greater generality in the sparse structure. Essentially, ICR solves a sequence of convex optimization problems such that sequence of solutions converges to a sub-optimal solution of the original hard optimization problem. Applications will be considered in image classification as well as image reconstruction.

Biography: Vishal Monga is a tenured Associate Professor in the School of Electrical Engineering and Computer Science at the Pennsylvania State University in University Park, PA. He was with Xerox Research from 2005-2009 and his doctoral work in Electrical Engineering was completed at the University of Texas, Austin in Aug 2005. His research interests are in computational imaging, statistical signal processing and convex optimization approaches to estimation problems. Prof. Monga is an elected member of the Editorial Board of the IEEE Transactions on Image Processing and the IEEE Signal Processing Letters. Prof. Monga is a recipient of the US National Science Foundation (NSF) CAREER award. Four of his papers have won best paper or Top 10 percent awards at IEEE Signal Processing conferences. He is a 2016 recipient of Joel and Ruth Spira Foundation Teaching Excellence award. He holds 40 US patents.