News

- In Spring 2015 I will be teaching a new course on high-dimensional inference and compressed sensing.
- Here is a video of a talk I gave in October 2013 at the Duke new faculty seminar: Robust compressed sensing: How undersampling introduces noise and what we can do about it
- I am currently looking for new graduate students. If you are already admitted at Duke please feel free to contact me. I am generally available to talk about research.
- I am currently looking or a postdoc with a strong background in signal processing, information theory, statistics, machine learning, or related field. If your are interested please contact me. Unfortunately, I will not be able to respond to all inquiries.

Research

My research interests lie at the intersection of signal processing, statistics, and information theory, with applications in compressed sensing, robust statistics, massive data storage and retrieval, neuroscience, and machine learning. With the era of information glut upon us, many of the most important scientific and technological advances of the next several decades will follow from our ability to collect, understand, and communicate massive amounts of data. An overall theme in my research is to draw upon mathematical tools from a wide variety of disciplines -- such as random matrix theory, convex optimization, statistical decision theory, and statistical physics -- to understand the limits of what is possible (and what is impossible) in problems of high-dimensional statistical inference, and also to figure out how we can reach these limits using computationally practical methods. My Ph.D. dissertation used tools from information theory to provide a sharp characterization of the problem of sparsity pattern recovery in compressed sensing.

Biography

Galen Reeves joined the faculty at Duke University in Fall 2013, and is currently an Assistant Professor with a joint appointment in the Department of Electrical & Computer Engineering and the Department of Statistical Science. He completed his PhD in Electrical Engineering and Computer Sciences at the University of California, Berkeley in 2011.
From 2011 to 2013 he was a postdoctoral associate in the Departments of Statistics at Stanford University, where he was supported by an NSF VIGRE fellowship. In the summer of 2011, he was a postdoctoral researcher in the School of Computer and Communication Sciences at EPFL, Switzerland; in the spring of 2009, he was a visiting scholar at the Technical University of Delft, The Netherlands;
and in the summer of 2008, he was a research intern in the Networked Embedded Computing Group at Microsoft Research, Redmond. He received his MS in Electrical Engineering from UC Berkeley in 2007, and BS in Electrical and Computer Engineering from Cornell University in 2005.

Publications (Updated November 2014)

Journal

*The Sampling Rate-Distortion Tradeoff for Sparsity Pattern Recovery in Compressed Sensing*G. Reeves and M. Gastpar, IEEE Transactions on Information Theory, vo. 58, no. 10, pp. 3065-3092, May, 2012. [arxiv]*Approximate Sparsity Pattern Recovery: Information-Theoretic Lower Bounds*G. Reeves and M. Gastpar, IEEE Transactions on Information Theory, vo. 59, no. 6, pp. 3451 - 3465, June, 2013. [arxiv]

Conference

*The Fundamental Limits of Stable Recovery in Compressed Sensing*G. Reeves, Proceedings of the IEEE International Symposium on Information Theory (ISIT 2014), Honolulu, HI, July 2014.*Beyond sparsity: Universally stable compressed sensing when the number of 'free' values is less than the number of observations.*G. Reeves, Proceedings of the 5th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, (CAMSAP 2013), Istanbul, Turkey, July 2013.*The Minimax Noise Sensitivity in Compressed Sensing*G. Reeves and D. L. Donoho, Proceedings of the IEEE International Symposium on Information Theory (ISIT 2013), Istanbul, Turkey, July 2013.*Achieving Bayes MMSE Performance in the Sparse Signal + Gaussian White Noise Model when the Noise Level is Unknown*D. L. Donoho and G. Reeves, Proceedings of the IEEE International Symposium on Information Theory (ISIT 2013), Istanbul, Turkey, July 2013.*The Sensitivity of Compressed Sensing Performance to Relaxation of Sparsity*D. L. Donoho and G. Reeves, Proceedings of the IEEE International Symposium on Information Theory (ISIT 2012), Boston, MA, July 2012.-
*Compressed Sensing Phase Transitions: Rigorous Bounds versus Replica Predictions*G. Reeves and M. Gastpar, Proceedings of the 46-th Annual Conference on Information Sciences and Systems (CISS 2012), Princeton, NJ, Mar 2012. *A Compressed Sensing Wire-Tap Channel*G. Reeves, N. Goela, N. Milosavljevic, and M. Gastpar, Proceedings of the IEEE Information Theory Workshop (ITW 2011), Paraty, Brazil, October 2011.*On the Role of Diversity in Sparsity Estimation*G. Reeves and M. Gastpar, Proceedings of the IEEE International Symposium on Information Theory (ISIT 2011), Saint Petersburg, Russia, August 2011. [See a short presentation on youtube]*"Compressed" Compressed Sensing*G. Reeves and M. Gastpar, Proceedings of the IEEE International Symposium on Information Theory (ISIT 2010), Austin, TX, June 2010.*A Note on Optimal Support Recovery in Compressed Sensing*G. Reeves and M. Gastpar, Proceedings of 43-rd Annual IEEE Asilomar Conference on Signals, Systems, and Computers, Monterey, CA, November 2009.-
*Managing Massive Time Series Streams with Multi-Scale Compressed Trickles*G. Reeves, J. Liu, S. Nath, and F. Zhao, Proceedings of the 35-th International Conference on Very Large Data Bases (VLDB 2009), Lyon, France, August 2009. *Efficient Sparsity Pattern Recovery*G. Reeves and M. Gastpar, Proceedings of the 30-th Symposium on Information Theory in the Benelux, Eindhoven, The Netherlands, May 2009.*Sampling Bounds for Sparse Support Recovery in the Presence of Noise*G. Reeves and M. Gastpar, Proceedings of the IEEE International Symposium on Information Theory (ISIT 2008), Toronto, Canada, July 2008.*Differences between Observation and Sampling Error in Sparse Signal Reconstruction*G. Reeves and M. Gastpar, Proceedings of the 2007 IEEE Workshop on Statistical Signal Processing (SSP 2007), Madison, Wisconsin, August, 2007.*Energy-Efficient Recursive Estimation by Variable Threshold Neurons*T. Berger, C. Levy, and G. Reeves, Presented at CoSyNe Workshop on Info-Neuro, Park City, UT, February, 2007.

Dissertation

*Sparsity Pattern Recovery in Compressed Sensing*G. Reeves, Ph.D. Thesis, Dec 2011.

Teaching

Courses at Duke

- STA 790.02 - Special Topics in High-Dimensional Statistical Inference, Spring 2015
- ECE 587 - Information Theory, Fall 2013.

Courses at Stanford

- STATS 110 - Statistical Methods in Engineering and the Physical Sciences, Autumn, 2011 and 2012.
- STATS 60 - Introduction to Statistical Methods: Precalculus, Spring, 2012.