BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//132.216.98.100//NONSGML kigkonsult.se iCalcreator 2.20.4//
BEGIN:VEVENT
UID:20260605T171620EDT-0181u5xM8S@132.216.98.100
DTSTAMP:20260605T211620Z
DESCRIPTION: \n\nThis seminar will feature Bogdan Mazoure\, who will tell u
 s about variational auto-encoders.\n\nAbstract:\n\nIn many statistical inf
 erence scenarios\, it is computationally inefficient or even impossible to
  calculate the true solution to a given problem. Instead\, we select a cla
 ss of approximate distributions (for example\, Normal) and minimize the di
 vergence between the true and approximate solutions through maximization o
 f a lower bound. Variational Bayesian methods are based on the free energy
  minimization principle used to explain physical systems and have a number
  of advantages over traditional inference models. In this talk\, we will g
 o over the main idea of approximate inference and discuss a neat use of va
 riational inference in deep learning via variational auto-encoders.\n
DTSTART:20180209T170000Z
DTEND:20180209T170000Z
LOCATION:Graduate Lounge\, BURN 1024/1025\, Burnside Hall\, CA\, QC\, Montr
 eal\, H3A 0B9\, 805 rue Sherbrooke Ouest
SUMMARY:Graduate Student Seminar - Bogdan Mazoure
URL:/mathstat/channels/event/graduate-student-seminar-
 bogdan-mazoure-284477
END:VEVENT
END:VCALENDAR