BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Learning Distributions for Physics: Sampling (FUND) and Inverse Pr
 oblems (EUSSIP)
DTSTART:20260910T090000Z
DTEND:20260910T100000Z
DTSTAMP:20260917T103300Z
UID:indico-event-9433@scitalks.tifr.res.in
DESCRIPTION:Speakers: Vipul Arora (ESAT\, KU Leuven)\n\nThis talk is about
  sampling and inverse problems in computational Physics.\nDeep generative 
 models (such as normalising flows) complement Markov chain Monte Carlo met
 hods for efficient sampling from high-dimensional distributions\, but they
  require training samples to perform well. I will talk about our recently 
 developed FUND algorithm\, which reliably trains a normalising flow model 
 without requiring any training samples. It obtains encouraging results for
  scalar phi^4 theory. \nThe second part of the talk is about simulation-b
 ased inverse problems. I will discuss our recent approach\, EUSSIP\, which
  employs uncertainty estimation along with active learning to iteratively 
 search for the inverse solution. Experiments show improved performance for
  inverse Laplace transform and LAE tomography in cosmology.\n \n\nhttps:/
 /scitalks.tifr.res.in/event/9433/
LOCATION:AG69 and On Zoom
URL:https://scitalks.tifr.res.in/event/9433/
END:VEVENT
END:VCALENDAR