BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Convergence Rate of Efficient MCMC with Ancillarity-Sufficiency In
 terweaving Strategy for Panel Data Models - Teruo Nakatsuma\, Keio Univers
 ity
DTSTART:20260312T150000Z
DTEND:20260312T160000Z
UID:TALK245059@talks.cam.ac.uk
CONTACT:Luning Sun
DESCRIPTION:Improving Markov chain Monte Carlo algorithm efficiency is ess
 ential for enhancing computational speed and inferential accuracy in Bayes
 ian analysis. These improvements can be effectively achieved using the anc
 illarity–sufficiency interweaving strategy (ASIS)\, an effective means o
 f achieving such gains. Herein\, we provide the first rigorous theoretical
  justification for applying ASIS in Bayesian hierarchical panel data model
 s. Asymptotic analysis demonstrated that when the product of prior varianc
 e of unobserved heterogeneity and cross-sectional sample size N is suffici
 ently large\, the latent individual effects can be sampled almost independ
 ently of their global mean. This near-independence accounts for ASIS’s r
 apid mixing behavior and highlights its suitability for modern “tall” 
 panel datasets. We derived simple inequalities to predict which convention
 al data augmentation scheme—sufficient augmentation (SA) or ancillary au
 gmentation (AA)—yields faster convergence. By interweaving SA and AA\, A
 SIS achieves optimal geometric rate of convergence and renders the Markov 
 chain for the global mean parameter asymptotically independent and identic
 ally distributed. Monte Carlo experiment confirm that this theoretical eff
 iciency ordering holds even for small panels (e.g.\, N = 10). These findin
 gs confirm the empirical success of ASIS application across finance\, mark
 eting\, and sports\, laying the groundwork for its extension to models wit
 h more complex covariate structures and non-Gaussian specifications.
LOCATION:S3.04\, Simon Sainsbury Centre\, Cambridge Judge Business School
END:VEVENT
END:VCALENDAR