BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Machine learning assisted accurate potential energy surfaces gener
 ation - Fabio E. A. Albertani
DTSTART:20201123T163000Z
DTEND:20201123T170000Z
UID:TALK154240@talks.cam.ac.uk
CONTACT:Bingqing Cheng
DESCRIPTION:Obtaining quantitatively accurate potential energy surfaces (P
 ESs) for molecular systems (with many dimensions and complex electronic st
 ructure) is challenging due to the increased computational cost related to
  simply getting the training data. It then becomes important to efficientl
 y select\, a priori\, the molecular geometries at which one wants to calcu
 late this expensive data\, and to use it efficiently.\n\nWe apply Gaussian
  processes to learn the energy surfaces as well as the corrections\, using
  delta-learning. We present various ways of using test sets from electroni
 c structure calculations\, as well as the intrinsic Gaussian processes cov
 ariance functions\, to generate optimal training sets for further PES gene
 ration. We present results for a water dimer proton exchange PESs from DFT
  data (PBE//aug-cc-pVDZ) to coupled cluster (CCSD(T)-F1\,2/aug-cc-pVTZ) ac
 curacy along all 12 dimensions. The same principles are also used on small
  molecules inside fullerenes at RIMP2//cc-pVDZ on the cage and RIMP2//cc-p
 VQZ on the internal molecule accuracies for the internal degrees of freedo
 m.
LOCATION:virtual ZOOM meeting ID: 263 591 6003\, Passcode: 000042\, https:
 //us02web.zoom.us/j/2635916003?pwd=ZlBEQnRENGwxNmJGMENGMWxjak5nUT09
END:VEVENT
END:VCALENDAR