Three-Toed Sloth

"Inference for Unlabelled Graphs" (This Week at the Statistics Seminar)

Attention conservation notice: Only of interest if you (1) care about the community discovery problem for networks and (2) will be in Pittsburgh on Friday.

I've talked about the community discovery problem here before, and even contributed to it; if you want a state-of-the-field you should read Aaron. This week, the CMU statistics seminar delivers a very distinguished statistician's take:

Peter Bickel, "Inference for Unlabelled Graphs"

Abstract:A great deal of attention has recently been paid to determining subcommunities on the basis of relations, corresponding to edges, between individuals, corresponding to vertices of an unlabelled graph (Newman, SIAM Review 2003; Airoldi et al, JMLR 2008; Leskovec, Kleinberg et al, SIGKDD 2005). We have developed a nonparametric framework for probabilistic ergodic models of infinite unlabelled graphs (PNAS 2009) and made some connections with modularities arising in the physics literature and community models in the social sciences. A fundamental difficulty in implementing these procedures is computational complexity. We show how a method of moments approach can partially bypass these difficulties.

This is joint work with Aiyou Chen and Liza Levina.

Place and time: Giant Eagle Auditorium, Baker Hall A51, 4:30--5:30 PM on Friday, April 23, 2010

As always, seminars are free and open to the public.

(This might motivate me to finally finish my post on Bickel and Chen's paper...)

Networks; Enigmas of Chance