BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//NextToU//EN
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:09f32ad7-e5b8-4811-859f-56ea332f5d98@nexttoyou
DTSTAMP:20261106T170609Z
DTSTART:20261113T170000Z
DTEND:20261113T180000Z
SUMMARY:Data Science and Statistics Seminar by Dr. Nicholas Marco (Duke)
LOCATION:Engineering and Computer Science South (ECSS) 2.203\, 800 W. Camp
 bell Road\, Richardson\, Texas 75080-3021
DESCRIPTION:Title: Adaptive Generalized Elliptical Slice Sampling\n\nAbstr
 act:\n\nA central goal in Bayesian computation is designing MCMC algorith
 ms that are broadly applicable\, bypass the need for manual tuning\, and 
 scale efficiently with dimension. Elliptical slice sampling meets many of
  these desiderata: it is tuning-free\, does not require differentiability
  of the target distribution\, automatically adapts its step size to local
  geometry\, and\, through generalized elliptical slice sampling\, can be 
 applied to a broad collection of target distributions. Under favorable co
 nditions\, elliptical slice sampling can achieve mixing times that scale 
 logarithmically with dimension\, making it more scalable than HMC and an 
 optimally tuned adaptive random walk. However\, this advantage is delicat
 e: its efficiency rapidly degrades if there is a mismatch between the tar
 get distribution and the distribution used to generate the ellipse-defini
 ng auxiliary variables\, explaining why we often observe poor efficiency 
 when using elliptical slice sampling in high-dimensional scenarios.\n\nTo
  overcome this substantial limitation\, we introduce adaptive generalized
  elliptical slice sampling (AGESS)\, which gradually adapts the distribut
 ion of the ellipse-defining auxiliary variables\, allowing the sampler to
  self-correct from a slow-mixing to a fast-mixing regime through online l
 earning. Under mild regularity conditions\, we prove that AGESS is ergodi
 c\, without requiring differentiability of the target and allowing for he
 \n\nhttp://next-to-u.app/events/09f32ad7-e5b8-4811-859f-56ea332f5d98
URL:http://next-to-u.app/events/09f32ad7-e5b8-4811-859f-56ea332f5d98
END:VEVENT
END:VCALENDAR