BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Network Representation Using Graph Root Distributions - Jing Lei\,
  Carnegie Mellon University
DTSTART:20200515T130000Z
DTEND:20200515T140000Z
UID:TALK140254@talks.cam.ac.uk
CONTACT:Dr Sergio Bacallado
DESCRIPTION:Exchangeable random graphs serve as an important probabilistic
  framework for the statistical analysis of network data.  This work introd
 uces an alternative parameterization for a large class of exchangeable ran
 dom graphs\, where the nodes are independent random vectors in a linear sp
 ace equipped with an indefinite inner product\, and the edge probability b
 etween two nodes equals the inner product of the corresponding node vector
 s. Therefore\, the distribution of exchangeable random graphs in this subc
 lass can be represented by a node sampling distribution on this linear spa
 ce\, which we call the "graph root distribution".  We study existence and 
 identifiability of such representations\, the topological relationship bet
 ween the graph root distribution and the exchangeable random graph samplin
 g distribution\, and estimation of graph root distributions.
LOCATION: https://zoom.us/j/95022384263?pwd=N3Z6elB2Vy9Jajd6azlCNjFHQVlKdz
 09
END:VEVENT
END:VCALENDAR