BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Stopping on the last success with unknown odds
DTSTART:20260402T090000Z
DTEND:20260402T101500Z
DTSTAMP:20260409T161100Z
UID:indico-event-15380@indico.math.cnrs.fr
DESCRIPTION:Speakers: Davy Paindaveine (Université Libre de Bruxelles)\n\
 nOptimal stopping problems lie at the interface of probability\, statistic
 s\, and sequential decision theory. They model situations in which decisio
 ns must be made online\, using only the information revealed so far and wi
 thout access to future outcomes. After briefly discussing a few motivating
  examples\, we will focus on the classical last-success problem: one obser
 ves a sequence of Bernoulli trials and seeks to stop exactly on the final 
 success. When the success probabilities are known\, the problem admits an 
 elegant and complete solution via Bruss’ sum-the-odds theorem (Bruss\, 2
 000\, Annals of Probability)\, which yields a simple threshold rule and an
  explicit formula for the optimal win probability. In most applications\, 
 however\, these probabilities are unknown\, so the oracle rule cannot be i
 mplemented. We therefore study the last-success problem under a minimal in
 formation structure\, where the decision maker observes only the sequentia
 l Bernoulli outcomes. Our contribution provides a quantitative theory of o
 racle-freeness: we analyze the performance of the natural plug-in odds rul
 e and\, more importantly\, characterize what is possible---and what is fun
 damentally impossible---when the success probabilities are unknown.\n\nhtt
 ps://indico.math.cnrs.fr/event/15380/
LOCATION:Auditorium 3 (Toulouse School of Economics)
URL:https://indico.math.cnrs.fr/event/15380/
END:VEVENT
END:VCALENDAR