BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Estimating network edge probabilities by neighborhood smoothing - 
 Elizaveta Levina (University of Michigan)
DTSTART:20160714T103000Z
DTEND:20160714T110000Z
UID:TALK66753@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:<span>Co-authors: Yuan Zhang (Ohio State University)\, Ji  Zhu
  (University of Michigan) <br></span> <br>The problem of estimating probab
 ilities of network edges from the observed  adjacency matrix has important
  applications to predicting missing links and  network denoising. It has u
 sually been addressed by estimating the graphon\, a  function that determi
 nes the matrix of edge probabilities\, but is ill-defined  without strong 
 assumptions on the network structure. Here we propose a novel  computation
 ally efficient method based on neighborhood smoothing to estimate the  exp
 ectation of the adjacency matrix directly\, without making the strong  str
 uctural assumptions graphon estimation requires. The neighborhood smoothin
 g  method requires little tuning\, has a competitive mean-squared error ra
 te\, and  outperforms many benchmark methods on the task of link predictio
 n in both  simulated and real networks. <br> <br>Related Links <ul> <li><a
  target="_blank" rel="nofollow">http://arxiv.org/abs/1509.08588</a>  - pap
 er&nbsp\;</li></ul>
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR