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:20210916T132528Z
LOCATION:Michel Mayor
DTSTART;TZID=Europe/Stockholm:20210708T170000
DTEND;TZID=Europe/Stockholm:20210708T190000
UID:submissions.pasc-conference.org_PASC21_sess167@linklings.com
SUMMARY:Data-Driven Scale-Bridging for Computational Materials Science
DESCRIPTION:Minisymposium\n\nThe field of computational materials design h
 as been pushed to new frontiers over the last decade. This is mainly roote
 d in recent advances to augment well known simulation approaches like dens
 ity functional theory (DFT) or molecular dynamics (MD) simulations with da
 ta-driven approaches like machine learning. The representation of DFT pote
 ntial energy surfaces with neural networks lead to a great gain in computa
 tional efficiency without losing much of the accuracy of the DFT calculati
 on. In addition to describing interatomic interactions, machine learning h
 as been further applied to predict material properties based on first-prin
 ciple calculations or to classify molecules based on thermodynamic propert
 ies. The ultimate goal of this new class of methodological approaches is t
 o aid the development of novel materials. These materials may find applica
 tions in drug design, unconventional energy resources or innovative semico
 nducting materials. This minisymposium will present state-of-the-art examp
 les of the underlying method development, scientific applications, and imp
 lementation on high performance computing platforms.\n\nIntegrated Machine
 -Learning Schemes for Atomic-Scale Materials Modeling\n\nCeriotti\n\nInter
 atomic potentials based on the statistical learning of energy and forces o
 btained from electronic-structure calculations have increased dramatically
  the time and length scales accessible to atomistic modeling. In this talk
  I will discuss how a new generation of machine-learning models is making.
 ..\n\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...\n\n----------
 -----------\nComputational High-Throughput Screening for Soft-Matter Mater
 ials\n\nBereau\n\nAdvanced statistical methods are rapidly impregnating ma
 ny scientific fields, offering new perspectives on long-standing problems.
  In materials science, data-driven methods are already bearing fruit in va
 rious disciplines, such as hard condensed matter or inorganic chemistry, w
 hile comparatively lit...\n\n---------------------\nActive Learning for Ma
 chine Learning Potentials\n\nSmith\n\nMachine learning (ML), trained on qu
 antum mechanics calculations, is a powerful tool for modeling potential en
 ergy surfaces of molecules and materials. The development of new neural ne
 twork architectures and intelligent data selection schemes has led to broa
 dly applicable ML-based potentials in rece...\n\n\nDomain: CS and Math, Ch
 emistry and Materials, Physics
END:VEVENT
END:VCALENDAR