BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:Canada/Eastern
BEGIN:DAYLIGHT
DTSTART:20220313T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:EDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20211107T010000
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:EST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20211201T002353Z
UID:396F10E9-B179-4335-8B1C-BB135688A4FD
DTSTART;TZID=Canada/Eastern:20211130T170000
DTEND;TZID=Canada/Eastern:20211130T183000
DESCRIPTION:Prerequisites:\nYou do not need to have attended the earlier ta
 lks. If you know zero math and zero machine learning\, then this talk is f
 or you. Jeff will do his best to explain fairly hard mathematics to you. I
 f you know a bunch of math and/or a bunch machine learning\, then these ta
 lks are for you. Jeff tries to spin the ideas in new ways.\nLonger Abstrac
 t:\nA randomly chosen bit string cannot be compressed at all. But if there
  is a pattern to it\, eg it represents an image\, then maybe it can be com
 pressed. Each pixel of an image is specified by one (or three) real number
 s. If an image has thousands/millions of pixels\, then each of these acts 
 as a coordinate of the point where the image sits in a very high dimension
 al space. A set of such images then corresponds to a set of\nthese points.
  We can understand the pattern of points/images as follows. Maximum Likeli
 hood assumes that the given set of points/images were randomly chosen acco
 rding a multi-dimensional normal distribution and then adjusts the paramet
 ers of this normal distribution in the way that maximizes the probability 
 of getting the images that we have. The obtained parameters effectively fi
 ts an ellipse around the points/images in this high dimensional space. We 
 then reduce the number of dimensions in our space by collapsing this ellip
 se along its least significant axises. Projecting each point/image to this
  lower dimensional space compresses the amount of information needed to re
 present each image.\n\nVirtual: https://events.vtools.ieee.org/m/289240
LOCATION:Virtual: https://events.vtools.ieee.org/m/289240
ORGANIZER:ayda.naserialiabadi@ryerson.ca
SEQUENCE:1
SUMMARY:Dimension Reduction & Maximum Likelihood: How to compress your data
  while retaining the key features
URL;VALUE=URI:https://events.vtools.ieee.org/m/289240
X-ALT-DESC:Description: <br /><p><strong>Prerequisites:</strong><br />You d
 o not need to have attended the earlier talks. If you know zero math and z
 ero machine learning\, then this talk is for you. Jeff will do his best to
  explain fairly hard mathematics to you. If you know a bunch of math and/o
 r a bunch machine learning\, then these talks are for you. Jeff tries to s
 pin the ideas in new ways.<br /><strong>Longer Abstract:</strong><br />A r
 andomly chosen bit string cannot be compressed at all. But if there is a p
 attern to it\, eg it represents an image\, then maybe it can be compressed
 . Each pixel of an image is specified by one (or three) real numbers. If a
 n image has thousands/millions of pixels\, then each of these acts as a co
 ordinate of the point where the image sits in a very high dimensional spac
 e. A set of such images then corresponds to a set of<br />these points. We
  can understand the pattern of points/images as follows. Maximum Likelihoo
 d assumes that the given set of points/images were randomly chosen accordi
 ng a multi-dimensional normal distribution and then adjusts the parameters
  of this normal distribution in the way that maximizes the probability of 
 getting the images that we have. The obtained parameters effectively fits 
 an ellipse around the points/images in this high dimensional space. We the
 n reduce the number of dimensions in our space by collapsing this ellipse 
 along its least significant axises. Projecting each point/image to this lo
 wer dimensional space compresses the amount of information needed to repre
 sent each image.</p>
END:VEVENT
END:VCALENDAR