New Tools for Distributed and Interactive Learning

In this talk we will describe an assortment of techniques and ideas for distributed and interactive learning algorithms. The topics will include use of codes for straggler-free large scale distributed gradient descent algorithms, interactive algorithms for clustering, and new statistical models for community detection. We will also provide new information theoretic lower bounds for these problems as well as near optimal and efficient algorithms. Time permitting, we will see how Shannon-type error-correcting codes exhibit some pseudorandom properties that make them excellent candidates for several sampling/counting problems such as compressed sensing, group testing, low-rank approximation, and computing partition functions.
Arya Mazumdar is an assistant professor in the College of Information and Computer Sciences at the University of Massachusetts Amherst since Fall 2015. Prior to this, Arya was an assistant professor at University of Minnesota-Twin Cities, and a postdoctoral scholar at Massachusetts Institute of Technology. Arya received his Ph.D. from University of Maryland, College Park, in 2011, where his thesis won a Distinguished Dissertation Fellowship Award. Arya is a recipient of the 2015 NSF CAREER award and the 2010 IEEE ISIT Jack K. Wolf Student Paper Award. He spent the summers of 2008 and 2010 at the Hewlett-Packard Laboratories, Palo Alto, CA, and IBM Almaden Research Center, San Jose, CA, respectively. Arya's research interests include coding theory, machine learning and information theory.