BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Europe/Warsaw
BEGIN:DAYLIGHT
DTSTART:20170326T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20171029T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20180208T222055Z
UID:F402CC3F-E5B6-11E7-833E-0050568D7F66
DTSTART;TZID=Europe/Warsaw:20170920T091500
DTEND;TZID=Europe/Warsaw:20170920T220000
DESCRIPTION:The introduction will reflect the view that powerful data model
 s are able to extract the true compositional nature of data\, which allows
  for a decomposition into their structural primitives.\nProbabilistic spar
 se coding and probabilistic version of non-negative matrix factorization (
 NMF) will be the first concrete models that are introduced and discussed. 
 The state-of-the-art of these models will then be used to point to recent 
 generalization directions. Two of these\, translation invariant versions a
 nd deep generalizations\, will be discussed in more detail. I will work ou
 t the benefits and challenges such novel approaches face\, and I will disc
 uss their crucial differences compared to supervised deep neural networks.
 \nFinally\, I briefly discuss semi-supervised approaches\, a field where m
 odern unsupervised and modern supervised Machine Learning algorithms come 
 together\, compete and where they are combined.\n\nSpeaker(s): Jörg Lück
 e\, \n\nPoznan\, Wielkopolskie\, Poland
LOCATION:Poznan\, Wielkopolskie\, Poland
ORGANIZER:tomasz.marciniak@put.poznan.pl
SEQUENCE:1
SUMMARY:[Legacy Report] Probabilistic and Unsupervised Machine Learning for
  Auditory Data and Pattern Recognition
URL;VALUE=URI:https://events.vtools.ieee.org/m/152549
X-ALT-DESC:Description: <br /><p>The introduction will reflect the view tha
 t powerful data models are able to extract the true compositional nature o
 f data\, which allows for a decomposition into their structural primitives
 .<br />Probabilistic sparse coding and probabilistic version of non-negati
 ve matrix factorization (NMF) will be the first concrete models that are i
 ntroduced and discussed. The state-of-the-art of these models will then be
  used to point to recent generalization directions. Two of these\, transla
 tion invariant versions and deep generalizations\, will be discussed in mo
 re detail. I will work out the benefits and challenges such novel approach
 es face\, and I will discuss their crucial differences compared to supervi
 sed deep neural networks.<br />Finally\, I briefly discuss semi-supervised
  approaches\, a field where modern unsupervised and modern supervised Mach
 ine Learning algorithms come together\, compete and where they are combine
 d.</p>
END:VEVENT
END:VCALENDAR