BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Scoring Gaussian process predictions for sequential design of expe
 riments - Lea Friedli (Technical University of Munich)
DTSTART:20250827T150000Z
DTEND:20250827T153000Z
UID:TALK234505@talks.cam.ac.uk
DESCRIPTION:Gaussian processes (GPs) have become a widely used tool for mo
 deling unknown functions across various domains. In many applications\, pa
 rticular interest lies in a specific range of the response\, with the goal
  of identifying inputs that lead to desired outputs. To enhance GP model p
 erformance in this setting\, we employ weighted scoring rules to develop s
 equential design strategies that selectively augment the training dataset.
  Specifically\, we study pointwise and integral criteria based on the thre
 shold-weighted Continuous Ranked Probability Score (CRPS)\, using two diff
 erent weighting measures. We showcase applications in synthetic chemistry\
 , where the objective is to identify molecules with specific properties\, 
 and in plant selection\, where the goal is to uncover combinations of geno
 types and environmental factors that yield desirable performances in wheat
 . However\, the presented acquisition strategies are applicable to a wide 
 range of fields and pave the way to further developing sequential design r
 elying on scoring rules.
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR