Imperial College London > Talks@ee.imperial > Featured talks > Approximate Message Passing and its application in low-complexity, rate-optimal communication
Log inImperial users Other users No account?Information onFinding a talk Adding a talk Syndicating talks Who we are Everything else |
Approximate Message Passing and its application in low-complexity, rate-optimal communicationAdd to your list(s) Download to your calendar using vCal
If you have a question about this talk, please contact Wei Dai. Abstract: “Approximate Message Passing” (AMP) refers to a class of iterative algorithms that are Gaussian or quadratic approximations of loopy belief propagation algorithms on dense factor graphs. AMP has attracted widespread interest as a technique for sparse signal recovery and related problems because of its fast convergence in many settings. (e.g., to solve the LASSO in compressed sensing reconstruction). In the first part of the talk, I will review the main ideas behind AMP , and give a result characterizing its finite sample performance. In the second part, I will describe an AMP decoding algorithm for sparse regression codes, a technique for communicating information over Gaussian noise channels. In this setting, the AMP decoder provably achieves the optimal information-theoretic limit (the channel capacity), and has excellent empirical performance as well. I will conclude with some open questions. This is joint work with Cynthia Rush and Adam Greig. The talk will be self-contained and will not assume prior knowledge of message passing. Bio: Ramji Venkataramanan is a University Lecturer in Information Engineering at the University of Cambridge where he is also a Fellow of Trinity Hall. He received Masters and Ph.D degrees in Electrical Engineering (Systems) from the University of Michigan, Ann Arbor in 2008, and his undergraduate degree from the Indian Institute of Technology, Madras in 2002. Before joining Cambridge in 2013, he held post-doctoral positions at Stanford University and Yale University. His research interests are broadly in the areas of information theory, statistical inference and learning. This talk is part of the Featured talks series. This talk is included in these lists:
Note that ex-directory lists are not shown. |
Other listsType the title of a new list here Talks@ee.imperial TalksOther talksModels and Capacity Bounds for Optical Fiber Channels (Information Theory Society Distinguished Lecture) Smart Urban Water Systems - A signal processing and communication perspective Holographic Radar and the Detection of micro-drone targets Perceptual Sound Field Recording, Reconstruction, and Synthesis |