Structured Signal Proc. & Learning

The goal of this project involves developing statistical signal processing methods in the context of structured information, a problem arising in many applications. Structure can be in the form of linear/non-linear/Boolean sensing operators, or in the form of signal structures such as sparsity, graph/group structures or low-rank information.

Selected Publications:

Sparse Signal Processing with Linear and Non-Linear Observations: A Unified Shannon Theoretic Approach (

Necessary and Sufficient Conditions and a Provably Efficient Algorithm for Separable Topic Discovery (

Learning Mixed Membership Mallows Models from Pairwise Comparisons (

Learning Immune-Defectives Graph through Group Tests (

Minimax Optimal Sparse Signal Recovery with Poisson Statistics (

Efficient Minimax Signal Detection on Graphs ( )