BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IC Monday Seminar : “A New Characterization of Compressed Sensin
 g Limits”
DTSTART:20110606T161500
DTSTAMP:20260528T015134Z
UID:b9b0e86a7bd2f4f49b75b7ad150b8f63b4d10bc3dce18c09351c192c
CATEGORIES:Conferences - Seminars
DESCRIPTION:Galen Reeves - UC Berkeley and EPFL invited by Prof. Michael G
 astpar\nAbstract: The fact that sparse vectors can be recovered from a sma
 ll number of linear measurements has important and exciting implications f
 or engineering and statistics. However\, despite the vast amount of recent
  work in the field of compressed sensing\, a sharp characterization betwee
 n what can and cannot be recovered in the presence of noise remains an ope
 n problem in general. In this talk\, we provide such a characterization fo
 r the task of sparsity pattern estimation (also known as support recovery)
 . Using tools from information theory\, we find a sharp separation into tw
 o problem regimes -- one in which the problem is fundamentally noise-limit
 ed\, and a more interesting one in which the problem is limited by the beh
 avior of the sparse components themselves. This analysis allows us to iden
 tify settings where existing computationally efficient algorithms\, such a
 s the LASSO\, are optimal as well as other settings where these algorithms
  are highly suboptimal. Furthermore\, we show how additional structure can
  make a key difference\, analogous to the role of diversity in wireless co
 mmunications. On the engineering side\, our analysis answers key engineeri
 ng questions related to compressed sensing: Is it better to increase SNR o
 r take more measurements? Is a given algorithm good enough? What accuracy 
 can be attained? On the mathematical side\, our results validate certain p
 hase transitions predicted by the powerful\, but heuristic\, replica metho
 d from statistical physics. Bio: Galen Reeves is currently a postdoctoral 
 researcher in the School of Computer and Communication Sciences\, EPFL\, w
 orking with Michael Gastpar. He received the B.S. degree in electrical and
  computer engineering from Cornell University in 2005 and the M.S. degree 
 in electrical engineering and computer sciences from the University of Cal
 ifornia at Berkeley (UC Berkeley) in 2007. His Ph.D. thesis is titled "A N
 ew Characterization of Compressed Sensing Limits\," and his research inter
 ests include statistical signal processing\, compressed sensing\, informat
 ion theory\, and machine learning. Beginning in Fall 2011\, he will be a V
 IGRE Postdoctoral Scholar in the Department of Statistics\, Stanford Unive
 rsity\, working with David Donoho.
LOCATION:INM 202
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR