[Colloq] machine learning talk at Math seminar, Wednesday September 11, 11 AM

[Rajmohan Rajaraman]

This talk on machine learning in the Math department should be of interest to several folks in CCIS.    

Best,

Rajmohan.

-------------------------------------------------

Applied and Interdisciplinary Mathematics Seminar 


Location: 509 Lake Hall, Northeastern University 
Special date: Wednesday, September 11 
Time: 11:00am - 12:00pm 

Anima Anandkumar 

UC Irvine 

Fast and Guaranteed Learning of Overlapping Communities via Tensor Methods 

Detecting hidden communities in networks is an important problem. A community refers to a group of related nodes. For instance, in a social network, it can represent individuals with shared interests or beliefs; in a gene network, it can represent genes with common regulatory mechanisms, and so on. Most previous approaches assume non-overlapping communities where a node can belong to at most one community. In contrast, we provide a guaranteed approach for detecting overlapping communities, when the network is generated from a class of probabilistic mixed membership models. Our approach is based on fast and scalable tensor decompositions and linear algebraic operations. We provide guaranteed recovery of community memberships and carry out a finite sample analysis of our algorithm. Additionally, our results match the lower bounds on scaling requirements (up to poly-log factors) in the special case of the stochastic block model (with non-overlapping communities). We have deployed the algorithm on GPUs, and our code design involves a careful optimization of GPU-CPU storage and communication. Our method is extremely fast and accurate on real datasets consisting of facebook network (about 20,000 nodes), yelp reviews (about 40,000 nodes) and dblp co-authorship network (about 120,000 nodes). For instance, on dblp dataset, our method takes under 2 hours to run to convergence. Thus, our approach is fast, scalable and accurate for detecting overlapping communities. 

Bio: Anima Anandkumar is a faculty at the EECS Dept. at U.C.Irvine. Her research interests are in the area of large-scale machine learning and high-dimensional statistics with a focus on learning probabilistic graphical models and latent variable models. She is the recipient of the Microsoft Faculty Fellowship, ARO Young Investigator Award, NSF CAREER Award, IBM Fran Allen PhD fellowship, and paper awards from the ACM SIGMETRICS and IEEE Signal Processing societies. She has been a visiting faculty at Microsoft Research New England and a postdoctoral researcher at the Stochastic Systems Group at MIT. She received her B.Tech in Electrical Engineering from IIT Madras and her PhD from Cornell University. 


http://www.math.neu.edu/content/applied-interdisciplinary-mathematics