Graph Signal Processing: Inferring Graphs from Data

Tuesday, April 24, 2018 - 3:00pm to 4:00pm

Event Calendar Category

LIDS Seminar Series

Speaker Name

Jose M F Moura

Affiliation

Carnegie Mellon University

Building and Room Number

32-141

Abstract

Signals of interest nowadays go well beyond time and image (space) signals to include corporate, health, social data, to mention a few application domains. Graphs may be used to explain dependencies among data and uncovering them is a major concern. I will briefly review the basics of Graph Signal Processing (Sandryhalia and Moura, 2013) and then present a methodology to infer a generic (directed, undirected, mixed) graph structure underlying multiple time series (Mei and Moura, 2017, 2018). Performance guarantees and applications with real data sets will be presented.

Biography

José M. F. Moura is the Philip L. and Marsha Dowd University Professor at CMU, with interests in signal processing and data science. He holds an EE degree from IST (Portugal) and a D.Sc. from MIT. He invented (with Alek Kavcic) a hard drive disk detector found in at least 60% of all computers sold worldwide in the last 14 years (over 3 billion and counting)–the subject of a $750 Million settlement between CMU and Marvell, the largest ever in the information technologies IP area and 3rd largest overall. He is the 2018 IEEE President-Elect. Moura received the IEEE SPS Technical Achievement Award and Society Award. He is Fellow of the IEEE, AAAS, US National Academy of Inventors, corresponding member of the Academy of Sciences of Portugal, and member of the US National Academy of Engineering.