BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Random sampling versus active learning algorithms for machine lear
 ning potentials of quantum liquid water - Dr. Nore Stolte\, Ruhr Universit
 y Bochum
DTSTART:20250203T143000Z
DTEND:20250203T150000Z
UID:TALK226453@talks.cam.ac.uk
CONTACT:Eszter Varga-Umbrich
DESCRIPTION:Today\, active learning is routinely used in training of machi
 ne learning potentials\, but the efficacy of active learning across differ
 ent systems is not well-tested. We have employed active learning algorithm
 s based on committee disagreement for the training of a high-dimensional n
 eural network potential for quantum liquid water at ambient conditions. No
 tably\, we compare active to random data selection\, given the same pool o
 f candidate structures. I will discuss in detail the performance of active
  learning\, at the level of computational requirements\, train and test er
 rors\, and the quality of the final potentials in simulations including nu
 clear quantum effects.\n
LOCATION:zoom.us/j/92447982065?pwd=RkhaYkM5VTZPZ3pYSHptUXlRSkppQT09
END:VEVENT
END:VCALENDAR