BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Machine learning potentials always extrapolate\, it does not matte
 r - Claudio Zeni\, SISSA\, Italy
DTSTART:20220131T140000Z
DTEND:20220131T143000Z
UID:TALK167267@talks.cam.ac.uk
CONTACT:Dr Christoph Schran
DESCRIPTION:Machine learning (ML) potentials for atomistic systems infer t
 he mapping between configurations and a target objective function\, e.g.\,
  the total energy of the system and/or the forces acting on each atom. We 
 show that\, contrary to popular assumptions\, predictions from machine lea
 rning potentials built upon atom-density representations almost exclusivel
 y occur in an extrapolation regime – i.e.\, in regions of the representa
 tion space which lie outside the convex hull defined by the training set p
 oints. We then propose a perspective to rationalise the domain of robust e
 xtrapolation and accurate prediction of atomistic machine learning potenti
 als in terms of the probability density induced by training points in the 
 representation space.
LOCATION:Venue to be confirmed
END:VEVENT
END:VCALENDAR