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:20260522T150116Z
LOCATION:D167/174
DTSTART;TZID=America/Chicago:20181111T153000
DTEND;TZID=America/Chicago:20181111T160000
UID:submissions.supercomputing.org_SC18_sess221_ws_mlhpce118@linklings.com
SUMMARY:Ramifications of Evolving Misbehaving Convolutional Neural Network
  Kernel and Batch Sizes
DESCRIPTION:Mark Coletti, Dalton Lunga, Anne Berres, Jibonananda Sanyal, a
 nd Amy Rose (Oak Ridge National Laboratory)\n\nDeep-learners have many hyp
 er-parameters including learning rate, batch size, kernel size --- all pla
 ying a significant role toward estimating high quality models.  Discoverin
 g useful hyper-parameter guidelines is an active area of research, though 
 the state of the art generally uses a brute force, uniform grid approach o
 r random search for finding ideal settings.  We share the preliminary resu
 lts of using an alternative approach to deep learner hyper-parameter tunin
 g that uses an evolutionary algorithm to improve the accuracy of a deep-le
 arner models used in satellite imagery building footprint detection. We fo
 und that the kernel and batch size hyper-parameters surprisingly differed 
 from sizes arrived at via a brute force uniform grid approach.  These diff
 erences suggest a novel role for evolutionary algorithms in determining th
 e number of convolution layers, as well as smaller batch sizes in improvin
 g deep-learner models.\n\nTag: Applications, Deep Learning, Machine Learni
 ng\n\nRegistration Category: Workshop Reg Pass\n\n
END:VEVENT
END:VCALENDAR