Brief Biography
I completed my B. Tech in Electrical
Engineering from IIT Bombay in 2000, after which I joined the University
of Illinois at Urbana-Champaign. There I obtained my MS in Electrical
Engineering in 2002, MS in Math in 2005, and PhD in Electrical Engineering
in 2006; my graduate research focused on communication network algorithms.
After obtaining my PhD, I joined LIDS, MIT as a postdoc. There I worked on
large-scale statistical inference and learning. In Fall 2008 I joined the
ECE department at Purdue as an assistant professor, and in Fall 2009 I
joined the ECE department at UT Austin.
My primary research interests are in probability and
optimization, as applied to the design and analyis of
algorithms in networks, communications and signal processing.
NEWS: Received an NSF CAREER Award for our research on Networks
and Statistical Inference (awarded January 2010)
Teaching:
Spring 2010: EE381V Sparsity,
Structure and Algorithms (graduate)
Fall 2009: EE381J Probability and
Stochastic Processes I (graduate)
Fall 2008 (in Purdue): ECE
302 Probabilistic Methods in Electrical and Computer Engineering (undergraduate)
Program Commitees: Infocom 2010, ICCCN 2009, MobiHoc 2009, WCNC 2009, WiCon 2008
Wireless and Sensor Networks
Routing, Scheduling, Gossip
Markov Random Fields, Graphical Models, Message
Passing
Statistical Inference and Learning
Game Theory
Graduate Students
Rajaganesh Ganesh
Ali Jalali
Praneeth Netrapalli
Recent Publications
Sparse and Low-Rank Matrix
Decompositions Venkat Chandrasekaran, Sujay Sanghavi, Pablo Parrilo,
Alan S. Willsky
Accepted to the 15th IFAC Sypmposium on System Identification
(SYSID), 2009
Submitted journal version:
Rank-Sparsity Incoherence for Matrix Decomposition
- Suppose we are given a matrix that is formed by adding an unknown sparse matrix to
an unknown low-rank matrix. This paper shows how to exactly decompose the given
matrix into its sparse and low-rank components, via convex programming. Based on a new
uncertainty principle between the rank and sparsity of a matrix.
Tightness of LP via Max-product Belief
Propagation Sujay Sanghavi, Devavrat Shah
Accepted to
Artificial Intelligence and Statistics (AISTATS) 2009.
- Shows that for certain random instances of the weighted independent set problem, the
(simple, edge-based) LP relaxation is tight with high probability. Uses belief propagation
as a proof technique.
Node Weighted Scheduling
Gagan Gupta, Sujay Sanghavi, Ness Shroff Sigmetrics 2009
- The standard approach to scheduling in input-queued switches determines schedules
based on the weights of edges. We develop an alternative approach that looks at the weights
of nodes instead. This gives new 100% throughput algorithms with lower complexity
and lower delay; proof involves a new Lyapunov function.
