BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Cambridge Psychometrics Centre Seminars
SUMMARY:Convergence Rate of Efficient MCMC with Ancillarit
 y-Sufficiency Interweaving Strategy for Panel Data
  Models - Teruo Nakatsuma\, Keio University
DTSTART;TZID=Europe/London:20260312T150000
DTEND;TZID=Europe/London:20260312T160000
UID:TALK245059AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/245059
DESCRIPTION:Improving Markov chain Monte Carlo algorithm effic
 iency is essential for enhancing computational spe
 ed and inferential accuracy in Bayesian analysis. 
 These improvements can be effectively achieved usi
 ng the ancillarity–sufficiency interweaving strate
 gy (ASIS)\, an effective means of achieving such g
 ains. Herein\, we provide the first rigorous theor
 etical justification for applying ASIS in Bayesian
  hierarchical panel data models. Asymptotic analys
 is demonstrated that when the product of prior var
 iance of unobserved heterogeneity and cross-sectio
 nal sample size N is sufficiently large\, the late
 nt individual effects can be sampled almost indepe
 ndently of their global mean. This near-independen
 ce accounts for ASIS’s rapid mixing behavior and h
 ighlights its suitability for modern “tall” panel 
 datasets. We derived simple inequalities to predic
 t which conventional data augmentation scheme—suff
 icient augmentation (SA) or ancillary augmentation
  (AA)—yields faster convergence. By interweaving S
 A and AA\, ASIS achieves optimal geometric rate of
  convergence and renders the Markov chain for the 
 global mean parameter asymptotically independent a
 nd identically distributed. Monte Carlo experiment
  confirm that this theoretical efficiency ordering
  holds even for small panels (e.g.\, N = 10). Thes
 e findings confirm the empirical success of ASIS a
 pplication across finance\, marketing\, and sports
 \, laying the groundwork for its extension to mode
 ls with more complex covariate structures and non-
 Gaussian specifications.
LOCATION:S3.04\, Simon Sainsbury Centre\, Cambridge Judge B
 usiness School
CONTACT:Luning Sun
END:VEVENT
END:VCALENDAR