On Friday September 16th Prof. Henry Pfister from Duke University had a DLT lecture at the Department of Electrical and Information Technology, Lund University. The topic of the lecture was “Graphical Models and Inference: Insights from Spatial Coupling”.
The slides of the lecture can be found here.
This talk focuses on recent theoretical and practical advances in coding, compressed sensing, and multiple-access communication based on spatially-coupled graphical models. The goal is to introduce the key ideas and insights using concrete examples. First, we introduce factor graphs and belief propagation (BP) as tools for understanding large systems of dependent random variables. Then, we describe how these techniques are applied to problems in signal processing and communications. Next, we use the example of low-density parity-check (LDPC) codes on the binary erasure channel to introduce the idea of density-evolution analysis. A key result is that BP decoding algorithms have a noise threshold below which recovery succeeds with high probability. Finally, we discuss how extrinsic-information transfer (EXIT) functions can be used to compare the performance between BP and optimal decoding.
Henry D. Pfister received his Ph.D. in electrical engineering in 2003 from the University of California, San Diego and he is currently an associate professor in the electrical and computer engineering department of Duke University. Prior to that, he was a professor at Texas A&M University (2006-2014), a post-doctoral fellow at the École Polytechnique Fédérale de Lausanne (2005-2006), and a senior engineer at Qualcomm Corporate R&D in San Diego (2003-2004).
He received the NSF Career Award in 2008, the Texas A&M ECE Department Outstanding Professor Award in 2010, the IEEE COMSOC best paper in Signal Processing and Coding for Data Storage in 2007, and a 2016 STOC Best Paper Award. He is currently an associate editor in coding theory for the IEEE Transactions on Information Theory (2013-2016) and a Distinguished Lecturer of the IEEE Information Theory Society (2015-2016).
His current research interests include information theory, communications, probabilistic graphical models, and machine learning.