BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Latent Variable Model\, Matrix Estimation and Collaborative Filter
 ing - Prof Devavrat Shah - MIT
DTSTART:20180509T151500Z
DTEND:20180509T161500Z
UID:TALK104260@talks.cam.ac.uk
CONTACT:David Greaves
DESCRIPTION:Estimating a matrix based on partial\, noisy observations is p
 revalent in variety of modern applications with recommendation system bein
 g a prototypical example. The non-parametric latent variable model provide
 s canonical representation for such matrix data when the underlying distri
 bution satisfies ``exchangeability’’ with graphons and stochastic bloc
 k model being recent examples of interest. Collaborative filtering has bee
 n a successfully utilized heuristic in practice since the dawn of e-commer
 ce. In this talk\, I will argue that collaborative filtering (and its vari
 ants) solve matrix estimationfor a generic latent variable model with near
  optimal sample complexity.\n\nThe talk is based on joint works with (a) C
 hristina Lee (MSR)\, Yihua Li (MS) and Dogyoon Song (MIT)\, and (b) Christ
 ina Borgs (MSR)\, Jennifer Chayes (MSR) and Christina Lee (MIT). 
LOCATION:Lecture Theatre 1\, Computer Laboratory
END:VEVENT
END:VCALENDAR