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:20230124T171520Z
LOCATION:C144-145
DTSTART;TZID=America/Chicago:20221114T153000
DTEND;TZID=America/Chicago:20221114T154500
UID:submissions.supercomputing.org_SC22_sess462_ws_ai4s105@linklings.com
SUMMARY:Ensuring AI For Science Is Science: Making Randomness Portable
DESCRIPTION:Workshop\n\nEnsuring AI For Science Is Science: Making Randomn
 ess Portable\n\nAhmed, Tchoua, Lofstead\n\nScience is a practice of system
 atically studying something and offering data and evidence to reach a conc
 lusion. With first principles simulations, basic physics are used to model
  some phenomena leading to consistent, repeatable results. With an incompl
 ete physics model or models too complex or costly to run for a given task,
  AI or ML are being used to estimate what the missing physics would be if 
 we could meet our goals with a first principles approach. Our work has bee
 n exploring how to ensure ML is capable of offering a science level of con
 sistency so we can trust our science applications incorporating ML models.
 \n\nOur earlier work examined the impact of pseudorandom numbers on model 
 quality. For this study, we have examined the pseudo-random number generat
 ion algorithms used to seed essentially all ML algorithms to ensure that m
 odel generation can be performed by other scientists to achieve identical 
 results.\n\nSession Format: Recorded\n\nRegistration Category: Workshop Re
 g Pass
END:VEVENT
END:VCALENDAR