DTC Seminar Series
Sample Complexity and Algorithms for Dictionary Learning from Tensor Data
Waheed U. Bajwa
Electrical and Computer Engineering
The State University of New Jersey
Friday, April 6, 2018
9:40 a.m. reception
10:00 a.m. seminar
101 Walter Library
Dictionary learning is a technique for finding sparse representations of data and, during the last decade, it has emerged as one of the most powerful methods for data-driven extraction of features from data. Although existing literature has mostly focused on dictionary learning for one-dimensional (i.e., vector-valued) data, many real-world data are multidimensional and have a tensor structure: examples include two-dimensional images, three-dimensional videos, and four-dimensional signals produced via magnetic resonance or computed tomography systems. In traditional dictionary learning literature, multidimensional data are converted into one-dimensional data by vectorization. Anecdotally, such approaches are known to result in poor sparse representations because they neglect the multidimensional structure of data. In this talk, we provide theoretical justification for this anecdotal evidence, which points to the need for development of dictionary learning algorithms that explicitly take into account the multidimensional structure of tensor data. In addition, we discuss a new algorithm, termed STARK, that has been explicitly designed for dictionary learning of tensor data.
(Sample complexity results are in collaboration with my PhD student Zahra Shakeri and colleague Anand Sarwate; STARK's development resulted from a collaboration with Zahra Shakeri, Mohsen Ghassemi, and Anand Sarwate.)
Waheed U. Bajwa received BE (with Honors) degree in electrical engineering from the National University of Sciences and Technology, Pakistan in 2001, and MS and PhD degrees in electrical engineering from the University of Wisconsin-Madison in 2005 and 2009, respectively. He was a Postdoctoral Research Associate in the Program in Applied and Computational Mathematics at Princeton University from 2009 to 2010, and a Research Scientist in the Department of Electrical and Computer Engineering at Duke University from 2010 to 2011. He is currently an Associate Professor in the Department of Electrical and Computer Engineering at Rutgers University. His research interests include statistical signal processing, high-dimensional statistics, machine learning, networked systems, and inverse problems.
Dr. Bajwa has received a number of awards in his career including the Best in Academics Gold Medal and President's Gold Medal in Electrical Engineering from the National University of Sciences and Technology (2001), the Morgridge Distinguished Graduate Fellowship from the University of Wisconsin-Madison (2003), the Army Research Office Young Investigator Award (2014), the National Science Foundation CAREER Award (2015), Rutgers University's Presidential Merit Award (2016), Rutgers Engineering Governing Council ECE Professor of the Year Award (2016, 2017), and Rutgers University's Presidential Fellowship for Teaching Excellence (2017). He is a co-investigator on the work that received the Cancer Institute of New Jersey's Gallo Award for Scientific Excellence in 2017, a co-author on papers that received Best Student Paper Awards at IEEE IVMSP 2016 and IEEE CAMSAP 2017 workshops, and a Member of the Class of 2015 National Academy of Engineering Frontiers of Engineering Education Symposium. He served as an Associate Editor of the IEEE Signal Processing Letters (2014 - 2017), co-guest edited a special issue of Elsevier Physical Communication Journal on "Compressive Sensing in Communications" (2012), co-chaired CPSWeek 2013 Workshop on Signal Processing Advances in Sensor Networks and IEEE GlobalSIP 2013 Symposium on New Sensing and Statistical Inference Methods, and served as the Publicity and Publications Chair of IEEE CAMSAP 2015 and General Chair of the 2017 DIMACS Workshop on Distributed Optimization, Information Processing, and Learning. He is currently serving as Technical Co-Chair of the IEEE SPAWC 2018 Workshop, Senior Area Editor for IEEE Signal Processing Letters, Associate Editor for IEEE Transactions on Signal and Information Processing over Networks, is a Senior Member of the IEEE, and serves on the MLSP, SAM, and SPCOM Technical Committees of the IEEE Signal Processing Society.
Fighting Covid 19 with Deep Graphs
Prof. George Karypis with PhD students Vasileios Ioannidis (ECE) and Saurav Manchanda (CSci) from the DTC at the University of Minnesota, a team of scientists from the Ohio State University, Hunan University, and Amazon’s AWS AI labs in Shanghai and Palo Alto have created the Drug Repurposing Knowledge Graph (DRKG) and a set of machine learning tools that can be used to prioritize drugs for repurposing studies. More on Fighting Covid 19
Digital Technology Center
499 Walter Library, 117 Pleasant Street SE, Minneapolis, MN 55455
P: (612) 624-9510