BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210916T132454Z
LOCATION:Lise Girardin
DTSTART;TZID=Europe/Stockholm:20210708T173000
DTEND;TZID=Europe/Stockholm:20210708T180000
UID:submissions.pasc-conference.org_PASC21_sess166_msa338@linklings.com
SUMMARY:Gradient-Based Methods for Black-Box Tuning and Optimisation
DESCRIPTION:Minisymposium\n\nGradient-Based Methods for Black-Box Tuning a
 nd Optimisation\n\nUstyuzhanin\n\nTuning simulation tools and finding opti
 mal device configurations is a challenging task that usually implies exper
 tise and significant computational investments. Well-known methods such as
  evolution algorithms or Bayesian optimisation help to address those chall
 enges. However, those approaches rely on assumptions that might not hold. 
 Recently, a series of methods have been introduced to estimate black-box g
 radients that significantly speed up the optimisation process. This talk o
 utlines such practices: REINFORCE-based, variational optimisation and surr
 ogate generative model-based approaches. We provide theoretical intuition 
 for those methods as well as practical illustrations of their strengths an
 d weaknesses. Such comparison will help practitioners to apply those metho
 ds to specific tasks in Cosmology and Particle Physics domains.\n\nDomain:
  CS and Math, Emerging Applications, Physics, Engineering
END:VEVENT
END:VCALENDAR