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:20210916T132453Z
LOCATION:Michel Mayor
DTSTART;TZID=Europe/Stockholm:20210708T170000
DTEND;TZID=Europe/Stockholm:20210708T173000
UID:submissions.pasc-conference.org_PASC21_sess167_msa310@linklings.com
SUMMARY:Enabling Predictive Scale-Bridging Simulations through Active Lear
 ning
DESCRIPTION:Minisymposium\n\nEnabling Predictive Scale-Bridging Simulation
 s through Active Learning\n\nRosenberger, Germann\n\nWe will describe effo
 rts at Los Alamos on active learning, which facilitates optimal training d
 ataset generation using uncertainty quantification built into the neural n
 etwork, and the computational challenges presented in deploying such metho
 ds on leadership-class computing platforms. These will be illustrated with
  three motivating applications: the performance of current machine learnin
 g-based potentials for soft materials; and transport through nanoporous ma
 terials and inertial confinement fusion (ICF), which both rely on accurate
  microscopic transport models that are provided by neural networks trained
  on molecular dynamics simulations. In particular, we have demonstrated a 
 proof of concept coupling between a kinetic code (Multi-BGK) with molecula
 r dynamics (LAMMPS) for an ICF application and the infrastructure / techni
 cal challenges involved. Ongoing work includes the efficient production of
  databases for transport properties and a general framework to learn an Eq
 uation of state. Both improve the generation of microscopic information, w
 hich facilitates more accurate bridging to larger scales.\n\nDomain: CS an
 d Math, Chemistry and Materials, Physics
END:VEVENT
END:VCALENDAR