BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Machine Learning Optimized Design of Experiments at the frontiers 
 of computation: methods and new perspectives - Pietro Vischia (Universidad
  de Oviedo)
DTSTART:20250626T091500Z
DTEND:20250626T094500Z
UID:TALK232288@talks.cam.ac.uk
DESCRIPTION:Designing the next generation colliders and detectors involves
  solving optimization problems in high-dimensional spaces where the optima
 l solutions may nest in regions that even a team of expert humans would no
 t explore. Furthermore\, the large amount of data we need to generate to s
 tudy physics for the next runs of large HEP machines and that we will need
  for future colliders is staggering\, requiring rethinking of our simulati
 on and reconstruction paradigm. Differentiable programming enables the inc
 orporation of domain knowledge\, encoded in simulation software\, into gra
 dient or reinforcement learning based pipelines\, resulting in the capabil
 ity of optimizing a given simulation setting and performing inference thro
 ugh classically intractable settings.\nIn this talk I will describe the fi
 rst proof-of-concept results for the gradient-based optimization of experi
 mental design\, with a focus on large-scale simulation software\, and will
  briefly touch on recent advances in calorimetry with neuromorphic hardwar
 e architectures\, paving the way to more complex challenges\, as well as o
 n the MODE Collaboration and the EUCAIF project.\n&nbsp\;
LOCATION:Seminar Room 1\, Newton Institute
END:VEVENT
END:VCALENDAR