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:6802@i2m.univ-amu.fr
DTSTART;TZID=Europe/Paris:20200217T140000
DTEND;TZID=Europe/Paris:20200217T170000
DTSTAMP:20241120T202013Z
URL:https://www.i2m.univ-amu.fr/evenements/marria-vladimirova-inria-grenob
 le-prior-specification-for-bayesian-deep-learning-models/
SUMMARY: (...): Mariia VLADIMIROVA (Inria Grenoble) - Prior specification f
 or Bayesian deep learning models
DESCRIPTION:: Mariia VLADIMIROVA (Inria Grenoble Rhone-Alpes in Mistis and 
 Thoth teams)\n\nTitle: \nPrior specification for Bayesian deep learning m
 odels\n \nAbstract: \nNeural networks (NNs)\, and their deep counterpart
 s\, have largely been used in many research areas such as image analysis\,
  signal processing\, or reinforcement learning\, just to name a few. The i
 mpressive performance provided by such machine learning approaches has gre
 atly motivated research that aims at a better understanding the driving me
 chanisms behind their effectiveness. In particular\, the study of the NNs 
 distributional properties through Bayesian analysis has recently gained mu
 ch attention.\nIn this seminar we firstly describe the necessary notations
  and statistical background for Bayesian NNs. Then we consider its distrib
 utional properties and novel theoretical insight on distributions at the u
 nits level. Under the assumption of independent and normally distributed w
 eights\, we establish that the induced prior distribution on the units bef
 ore and after activation becomes increasingly heavy-tailed with the depth 
 of the layer. Lastly\, we discuss this property in terms of a regularizing
  mechanism and corroborate it with experimental simulation results.\n\n
ATTACH;FMTTYPE=image/jpeg:https://www.i2m.univ-amu.fr/wp-content/uploads/2
 020/02/Mariia-Vladimirova.jpg
CATEGORIES:Séminaire,Statistique
LOCATION:Saint-Charles - FRUMAM  (2ème étage)\, 3 Place Victor Hugo\, Mar
 seille\, 13003\, France
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=3 Place Victor Hugo\, Marse
 ille\, 13003\, France;X-APPLE-RADIUS=100;X-TITLE=Saint-Charles - FRUMAM  (
 2ème étage):geo:0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:STANDARD
DTSTART:20191027T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
END:STANDARD
END:VTIMEZONE
END:VCALENDAR