University of Minnesota
University Relations
http://www.umn.edu/urelate
612-624-6868
myU OneStop


Go to unit's home.

Home | Seminars and Symposia | Past seminars/symposia: Thursday, April 28, 2016

DTC Seminar Series

Distributed Stochastic Convex Optimization

by

Michael Rabbat
McGill University
Montreal, Canada

Thursday, April 28, 2016
4:00 p.m. reception
4:30 p.m. seminar

101 Walter Library

Michael Rabbat

This talk considers the problem of distributed convex optimization in a stochastic setting. Each node in a network of processors has a stochastic oracle for a common objective function, and the aim of the network is to collectively minimize the objective as quickly as possible. Such a problem arises, for example, in large-scale machine learning where the goal of the network is to fit a model to training data that is stored at multiple nodes. We study a consensus-based approach where nodes individually take descent steps and then consensus iterations are performed to synchronize models across the nodes. We prove that the proposed method achieves the optimal centralized regret bound when the objective function has Lipschitz continuous gradients, and we discuss the tradeoff between communication, computation, and the network topology. This is joint work with Konstantinos Tsianos.

 

Michael Rabbat earned the B.Sc. degree from the University of Illinois, Urbana-Champaign, in 2001, the M.Sc. degree from Rice University in 2003, and the Ph.D. from the University of Wisconsin, Madison, in 2006, all in electrical engineering. He joined McGill University, Montreal, Canada, in 2007, and he is currently an Associate Professor. During the 2013--2014 academic year he held visiting positions at Telecom Bretegne, Brest, France, the Inria Bretagne-Atlantique Reserch Centre, Rennes, France, and KTH Royal Institute of Technology, Stockholm, Sweden. He was a Visiting Researcher at Applied Signal Technology, Inc., Sunnyvale, USA, during the summer of 2003. Dr. Rabbat co-authored the paper which received the Best Paper Award (Signal Processing and Information Theory Track) at the 2010 IEEE International Conference on Distributed Computing in Sensor Systems (DCOSS). He received an Honorable Mention for Outstanding Student Paper Award at the 2006 Conference on Neural Information Processing Systems (NIPS) and a Best Student Paper Award at the 2004 ACM/IEEE International Symposium on Information Processing in Sensor Networks (IPSN). He currently serves as Senior Area Editor for IEEE Signal Processing Letters and as Associate Editor for IEEE Transactions on Signal and Information Processing over Networks and IEEE Transactions on Control of Network Systems. His research interests include graph signal processing, distributed algorithms for optimization and inference, consensus algorithms, and network modelling and analysis, with applications in distributed sensor systems, large-scale machine learning, statistical signal processing, and social networks.