[Dmbu-l] Fwd: FW: Probability and Statistics Seminar at BU for next week

George Kollios gkollios at bu.edu
Fri Sep 20 14:45:35 EDT 2013

---------- Forwarded message ----------
From: Crovella, Mark E <crovella at bu.edu>
Date: Fri, Sep 20, 2013 at 12:55 PM
Subject: FW: Probability and Statistics Seminar at BU for next week
To: "Terzi, Evimaria D" <evimaria at bu.edu>, "Kollios, George" <
gkollios at bu.edu>

  From: Konstantinos Spiliopoulos <kspiliop at bu.edu>
Reply-To: "Spiliopoulos, Konstantinos" <kspiliop at bu.edu>
Date: Friday, September 20, 2013 8:44 AM
To: "prob-sem at math.bu.edu" <prob-sem at math.bu.edu>, "stat-seminar at math.bu.edu"
<stat-seminar at math.bu.edu>
Subject: Probability and Statistics Seminar at BU for next week

  Dear all,

This is an announcement   for the statistics and probability seminar
at BU for the coming week, on Thursday,  September the 26th.

Details are below:

Location: 111 Cummington Mall, Department of Mathematics and
Statistics, BU, 02215,

*Thursday 4:00-5:00pm, Room MCS 148
(Tea served from 3:30-4:00pm, Room MCS 144)*

Speaker:  *Jiashun Jin, Department of Statistics, Carnegie Mellon
University <http://www.stat.cmu.edu/%7Ejiashun/>, Thursday 26 September 2013

 Title:  Fast Network Community Detection by SCORE

Abstract:   Consider a network where the nodes split into K different
communities. The community labels for the nodes are unknown and it is of
major interest to estimate them (i.e., community detection). Degree
Corrected Block Model (DCBM) is a popular network model. How to detect
communities with the DCBM is an interesting problem, where the main
challenge lies in the degree heterogeneity.

We propose Spectral Clustering On Ratios-of-Eigenvectors (SCORE) as a new
approach to community detection. Compared to existing spectral methods, the
main innovation is to use the entry-wise ratios between the first a few
leading eigenvectors for community detection. The central surprise is, the
effect of degree heterogeneity is largely ancillary, and can be effectively
removed by taking such entry-wise ratios. We have applied SCORE to the
well-known web blogs data and the statistics co-author network data which
we have collected very recently. We find that SCORE is competitive both in
computation and in performance. On top of that, SCORE is conceptually
simple and has the potential for extensions in various directions. Addi-
tionally, we have identi ed several interesting communities in
statisticians, including what we call the \Object Bayesian community",
\Theoretic Machine Learning Com- munity", and the \Dimension Reduction

We develop a theoretic framework where we show that under mild regularity
conditions, SCORE stably yields consistent community detection. In the core
of the analysis is the recent development on Random Matrix Theory (RMT),
where the matrix-form Bernstein inequality is especially helpful.

with regards,

Konstantinos Spiliopoulos
Assistant Professor
Boston University
Department of Mathematics & Statistics
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://cs-mailman.bu.edu/pipermail/dmbu-l/attachments/20130920/59fcc6b7/attachment.html>

More information about the Dmbu-l mailing list