BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20190719T085743Z
LOCATION:HG F 1
DTSTART;TZID=Europe/Stockholm:20190612T163000
DTEND;TZID=Europe/Stockholm:20190612T170000
UID:submissions.pasc-conference.org_PASC19_sess145_msa239@linklings.com
SUMMARY:Exploiting Sparsity in the Estimation of Gaussian Models at Large 
 Scales
DESCRIPTION:Minisymposium\nComputer Science and Applied Mathematics, Chemi
 stry and Materials, Physics, Engineering\n\nExploiting Sparsity in the Est
 imation of Gaussian Models at Large Scales\n\nTreister\n\nThe Gaussian dis
 tribution is one of the most fundamental statistical tools for modeling da
 ta in various applications. However, estimating full covariance matrices i
 s both prohibitively expensive and over-parametrized at large scales. In t
 his talk I will discuss how to exploit sparsity to overcome both issues, a
 nd present two applications. To start the talk I will present a multilevel
  acceleration for estimating of sparse precision matrices at large scales.
  Then I will discuss two applications. One application targets image proce
 ssing tasks using sparse Gaussian mixtures for modeling of image
  and video patches. The other application targets modeling of sparse singl
 e cell genomic data, and uses the Gaussian distribution for the imputing t
 he data in an expectation-maximization framework.
END:VEVENT
END:VCALENDAR