BEGIN:VCALENDAR
VERSION:2.0
PRODID:researchseminars.org
CALSCALE:GREGORIAN
X-WR-CALNAME:researchseminars.org
BEGIN:VEVENT
SUMMARY:Charles-Edouard Bréhier (Université de Pau et des Pays de l'Adou
 r)
DTSTART:20241007T111500Z
DTEND:20241007T120000Z
DTSTAMP:20260417T004254Z
UID:cam/35
DESCRIPTION:Title: <a href="https://researchseminars.org/talk/cam/35/">Asy
 mptotic error analysis of stochastic optimization schemes</a>\nby Charles-
 Edouard Bréhier (Université de Pau et des Pays de l'Adour) as part of CA
 M seminar\n\nLecture held in MV:L14.\n\nAbstract\nStochastic optimization 
 algorithms are nowadays widely used\, especially in the machine learning c
 ommunity. In this talk\, we study a class of stochastic optimization schem
 es which are perturbations of gradient descent algorithms.\nWe perform a r
 igorous analysis of the convergence\, with proofs of error bounds with res
 pect to the time-step size\, in the large time regime\, in the case of str
 ongly convex objective functions. The error bounds follow from an interpre
 tation of the schemes in terms of deterministic and stochastic modified eq
 uations\, and using tools from weak error analysis of numerical methods fo
 r stochastic differential equations.\n
LOCATION:https://researchseminars.org/talk/cam/35/
END:VEVENT
END:VCALENDAR