BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20260329T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251026T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260424T114113Z
UID:68ff2edf36c8d703287353@ist.ac.at
DTSTART:20251105T110000
DTEND:20251105T120000
DESCRIPTION:Speaker: Giovanni Conforti\nhosted by Filippo Quattrocchi / Jan
  Maas\nAbstract: In this talk\, I will survey some recent progresses in th
 e field of entropic optimal transport (a.k.a. Schrdinger problem). More pr
 ecisely I will focus on results providing theoretical guarantees of conver
 gence for the most popular algorithms employed to compute approximate solu
 tions in machine learning applications\, namely Sinkhorns algorithm and th
 e iterated Markovian fitting algorithm. In particular\, I will develop the
  connection between smoothness of entropic potentials\, stability of optim
 al solutions and exponential convergence in the number of iterations. The 
 theoretical results will be illustrated in some concrete examples\, such a
 s log-concave marginals and marginals with bounded support\, where the exp
 onential rates are shown to have an optimal dependence on the regularizati
 on parameter.Based on joint work with A.Chiarini\, A.Durmus\, M.Gentiloni\
 , G.Greco\, L.Tamanini.
LOCATION:Sunstone Bldg / Ground floor / Big Seminar Room A (I23.EG.102)\, I
 STA
ORGANIZER:boosthui@ist.ac.at
SUMMARY:Giovanni Conforti: Theoretical guarantees of convergence for Sinkho
 rn’s algorithm and the iterative Markovian Fitting procedure
URL:https://talks-calendar.ista.ac.at/events/6098
END:VEVENT
END:VCALENDAR