BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Michael Tretyakov (University of Nottingham)
DTSTART:20260413T111500Z
DTEND:20260413T120000Z
DTSTAMP:20260417T003019Z
UID:cam/94
DESCRIPTION:Title: <a href="https://researchseminars.org/talk/cam/94/">Neu
 ral SDEs for variance reduction</a>\nby Michael Tretyakov (University of N
 ottingham) as part of CAM seminar\n\nLecture held in MV:L14.\n\nAbstract\n
 Variance reduction techniques are of crucial importance for increasing eff
 iciency of Monte Carlo simulations. Neural stochastic differential equatio
 ns (SDEs)\, with control variates parameterized by neural networks\, are c
 onsidered in order to learn approximately optimal control variates and hen
 ce reduce variance. A black-box fashion practical variance reduction tool\
 , which does not require any lengthy pre-training and tuning\, is proposed
  for both SDEs driven by Brownian motion and\, more generally\, by Lévy p
 rocesses including those with infinite activity. For the latter case\, opt
 imality conditions for the variance reduction are proved. Weak approximati
 on of SDEs governed by infinite-activity Lévy processes is also discussed
 . Several numerical examples from option pricing are presented. The talk i
 s mainly based on a joint work with Piers Hinds.\n
LOCATION:https://researchseminars.org/talk/cam/94/
END:VEVENT
END:VCALENDAR