BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:177@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20140321T140000
DTEND;TZID=Europe/Paris:20140321T150000
DTSTAMP:20140306T130000Z
URL:https://www.i2m.univ-amu.fr/evenements/learning-negative-mixture-model
 s-by-tensor-decomposition/
SUMMARY: (...): Learning negative mixture models by tensor decomposition
DESCRIPTION:: In this talk\,  we consider the problem of estimating the par
 ameters of negative mixture models\, i.e. mixture models that possibly inv
 olve negative weights. We show that every rational probability distributio
 ns on strings\, a representation which occurs naturally in spectral learni
 ng\, can be computed by a negative mixture of at most two  probabilistic a
 utomata (or HMMs).  We present a method to estimate the parameters of nega
 tive mixture models having a specific tensor structure in their low order 
 observable moments. Building upon a recent paper on tensor decompositions 
 for learning latent variable models\, we extend this work to the broader s
 etting of tensors having a symmetric decomposition with positive and negat
 ive weights. This extension leads to a generalisation of the tensor power 
 method for complex valued tensors\, for which we establish theoretical con
 vergence guarantees. Finally\, we show how our approach applies to negativ
 e Gaussian mixture models.http://www.lif.univ-mrs.fr/annuaire/personne/122
 9875
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:STANDARD
DTSTART:20131027T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
END:VTIMEZONE
END:VCALENDAR