BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20181221T160904Z
LOCATION:C2/3/4 Ballroom
DTSTART;TZID=America/Chicago:20181113T083000
DTEND;TZID=America/Chicago:20181113T170000
UID:submissions.supercomputing.org_SC18_sess322_post130@linklings.com
SUMMARY:Cross-Layer Group Regularization for Deep Neural Network Pruning
DESCRIPTION:Poster\nTech Program Reg Pass, Exhibits Reg Pass\n\nCross-Laye
 r Group Regularization for Deep Neural Network Pruning\n\nGao, Liu\n\nImpr
 oving weights sparsity is a common strategy for deep neural network prunin
 g. Most existing methods use regularizations that only consider structural
  sparsity within an individual layer. In this paper, we propose a cross-la
 yer group regularization taking into account the statistics from multiple 
 layers. For residual networks, we use this approach to align kernel sparsi
 ty across layers that are tied to each other through element-wise operatio
 ns: the ith kernel of these layers are put into one regularization group, 
 they either stay or be removed simultaneously during pruning. In this way,
  the computational and parameter storage cost could be significantly reduc
 ed. Experimental results show that this method does not only  improve weig
 hts sparsity but also align kernel weights sparsity across related layers.
  Our method is able to prune ResNet up to 90.4% of parameters and improve 
 runtime by 1.5x speedup, without loss of accuracy.
URL:https://sc18.supercomputing.org/presentation/?id=post130&sess=sess322
END:VEVENT
END:VCALENDAR