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:20230124T171521Z
LOCATION:C1-2-3
DTSTART;TZID=America/Chicago:20221117T083000
DTEND;TZID=America/Chicago:20221117T170000
UID:submissions.supercomputing.org_SC22_sess275_rpost124@linklings.com
SUMMARY:Distributed Deep Learning on HPC for Infilling Holes in Spatial Pr
 ecipitation Data
DESCRIPTION:Posters, Research Posters\n\nDistributed Deep Learning on HPC 
 for Infilling Holes in Spatial Precipitation Data\n\nMeuer, Plésiat, Thiem
 ann, Ludwig, Kadow\n\nMissing climatological data is a general problem in 
 climate research that leads to uncertainty of prediction models that rely 
 on these data resources. So far, existing approaches for infilling missing
  precipitation data are mostly numerical or statistical techniques that re
 quire time consuming computations and are not suitable for large regions w
 ith missing data. Most recent machine learning techniques have proven to p
 erform well on infilling missing temperature or satellite data. However, t
 hese techniques consider only spatial variability in the data whereas prec
 ipitation data is much more variable in both space and time. We propose a 
 convolutional inpainting network that additionally considers temporal vari
 ability and atmospheric parameters in the data. The model was trained and 
 evaluated on the RADOLAN data set over Germany. Since the training of this
  high-resolved data set requires a large amount of computational resources
 , we apply distributed training on an HPC system to maximize the performan
 ce.\n\nRegistration Category: Tech Program Reg Pass, Exhibits Reg Pass
END:VEVENT
END:VCALENDAR