BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:NUTS for NUTS: New Advances in No-U-Turn Samplers [Colloquium CRC 
 1720]
DTSTART:20260609T121500Z
DTEND:20260609T131500Z
DTSTAMP:20260722T042600Z
UID:indico-event-956@math-events.uni-bonn.de
CONTACT:sfb1720-seminar@uni-bonn.de
DESCRIPTION:Speakers: Nawaf Bou-Rabee\n\nMarkov chain Monte Carlo (MCMC) r
 emains a central tool for sampling from intractable distributions\, yet th
 e efficiency of classical algorithms often deteriorates in high dimension
 s or anisotropic geometries. The No-U-Turn Sampler (NUTS) and its descend
 ants have transformed practical Bayesian computation by adapting trajector
 y lengths to local geometry\, enabling efficient exploration even in comp
 lex\, high-dimensional landscapes.  Despite their empirical success\, a r
 igorous understanding of why such locally adaptive schemes mix efficiently
  has remained elusive.\nThis talk revisits the mathematical foundations of
  NUTS and shows how they can be extended and unified within a broader fram
 ework. This perspective leads to new algorithms that preserve the self-tun
 ing spirit of NUTS while extending its reach to ill-conditioned geometries
 .  Along the way\, we will see how No-U-Turn ideas are evolving from clev
 er computational innovations into a principled theory of locally adaptive 
 MCMC\, bringing us closer to the long-standing program of constructing sam
 plers that require no tuning with provable efficiency guarantees.\n\nhttps
 ://math-events.uni-bonn.de/event/956/
LOCATION:Endenicher Allee 60/1-016 - Lipschitzsaal (Mathezentrum)
URL:https://math-events.uni-bonn.de/event/956/
END:VEVENT
END:VCALENDAR