BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Ranking algorithms on directed configuration networks - Nelly Litv
 ak (Universiteit Twente)
DTSTART:20160715T083000Z
DTEND:20160715T090000Z
UID:TALK66770@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:<span>Co-authors: Ningyuan Chen (Yale School of  Management)\,
  Mariana Olvera-Cravioto (Columbia University) <br></span> <span><br>We st
 udy a family of rankings\, which includes Google&#39\;s PageRank\, on a  d
 irected configuration model. We show that the the rank of a randomly chose
 n  vertex converges in distribution to a finite random variable that can b
 e written  as a linear combination of i.i.d. copies of the attracting endo
 genous solution  to a stochastic fixed-point equation. We provide precise 
 asymptotics for this  limiting random variable. In particular\, if the in-
 degree distribution in the  directed configuration model has a power law d
 istribution\, then the limiting  distribution of the rank also follows a p
 ower law with the same exponent. Such  power law behaviour of ranking is w
 ell-known from empirical studies of real-life  networks. Our asymptotic re
 sult gives remarkably good approximation for the  complete ranking distrib
 ution on configuration networks of moderate size and on  the directed grap
 h of English Wikipedia.</span>
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR