BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Convex low-rank models: from matrices to tensors - Ryota Tomioka\,
  Toyota Technological Institute at Chicago
DTSTART:20150310T133000Z
DTEND:20150310T143000Z
UID:TALK58379@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:In this talk I will present low-rank models in two domains and
  how they can be set-up as convex optimization problems. The fist domain i
 s the so-called brain-computer interface. The problem of learning a set of
  discriminative spatio-temporal filters for the P300 speller system is for
 mulated as a low-rank matrix learning problem. Using the trace norm regula
 rization combined with an appropriately defined likelihood function I obta
 ined both state-of-the art classification performance and highly interpret
 able spatio-temporal filters. The second domain is audio-signal separation
 . I propose positive semidefinite tensor factorization (PSDTF)\, which is 
 a generalization of nonnegative matrix factorization (NMF)\, that decompos
 es a collection of PSD matrices into linear combinations of a small number
  of basis PSD matrices. I propose a non-parametric Bayesian model for PSDT
 F that automatically infers the number of basis PSD matrices. Finally\, I 
 present an ongoing work on convex relaxation of tensor (multilinear)-rank.
  I propose generalizations of trace norm for tensors and analyze their sta
 tistical performance. The wide gap between the performance that can be obt
 ained by tractable algorithms and that can only be obtained by intractable
  algorithms points to an optimization-statistics tradeoff that opens up ma
 ny future direction.\nIncludes joint work with Klaus-Robert Müller\, Kazu
 yoshi Yoshii\, Daichi Mochihashi\, Masataka Goto\, Taiji Suzuki\, Kohei Ha
 yashi\, and Hisashi Kashima.
LOCATION:Auditorium\, Microsoft Research Ltd\, 21 Station Road\, Cambridge
 \, CB1 2FB
END:VEVENT
END:VCALENDAR