DIMACS Workshop on Markov Chain Monte Carlo: Synthesizing Theory and Practice

June 4 - 7, 2007
DIMACS Center, CoRE Building, Rutgers University

Organizers:

Jim Fill, Johns Hopkins University, jimfill@jhu.edu

Jim Hobert, University of Florida, jhobert@stat.ufl.edu

Antonietta Mira, University of Insubria (Italy), amira@eco.unisubria.it

Luke Tierney, University of Iowa, luke@stat.uiowa.edu

Peter Winkler, Dartmouth College, peter.winkler@dartmouth.edu

Workshop Program:

Monday, June 4, 2007

 8:45 -  9:20 Breakfast and Registration

 9:20 -  9:30 Welcoming Remarks
              Tami Carpenter, DIMACS Associate Director

 9:30 - 10:15 Tutorial: Markov Chain Monte Carlo: Theory, Part I
              Eric Vigoda, Georgia Institute of Technology

10:15 - 10:45 Break

10:45 - 11:30 Tutorial: Markov Chain Monte Carlo: Theory, Part II
              Eric Vigoda, Georgia Institute of Technology
 
11:40 - 12:25 Tutorial: Markov Chain Monte Carlo: Practice;
              or: The Full Monte Carlo: A Live Performance (with Stars), Part I
              Xiao-Li Meng, Harvard University

12:30 -  2:30 Lunch

 2:30 -  3:15 Tutorial: Markov Chain Monte Carlo: Practice;
              or: The Full Monte Carlo: A Live Performance (With Stars), Part II
              Xiao-Li Meng, Harvard University

 3:25 -  4:10 Tutorial: Markov Chain Monte Carlo: Synthesizing Theory and Practice, Part I
              Jeffrey Rosenthal, University of Toronto (Canada)

 4:10 -  4:40 Break

 4:40 -  5:25 Tutorial: Markov Chain Monte Carlo: Synthesizing Theory and Practice, Part II
              Jeffrey Rosenthal, University of Toronto (Canada)

Tuesday, June 5, 2007

 8:45 -  9:30 Breakfast and Registration

 9:30 - 10:15 Binary Contingency Tables in Theory and Practice
              Ivona Bezakova, Rochester Institute of Technology

10:15 - 10:45 Break

10:45 - 11:30 MCMC Algorithms for Distributions with Intractable Normalizing
              Constants, with a View to Perfect Simulation and Non-parametric
              Bayesian Inference for Inhomogeneous Markov Point Processes
              Jesper Møller, Aalborg University (Denmark)
 
11:40 - 12:25 Computing for Bayesian Spatial Estimation and Prediction 
              with Application to Residential Radon
              Kate Cowles, University of Iowa

12:30 -  2:30 Lunch

 2:30 -  3:15 Designing and Testing Efficient Sequential Importance Sampling Algorithms
              Jose Blanchet, Harvard University
           
 3:25 -  4:10 Using SIS to Speed Up MCMC
              Isabel M. Beichl, National Institute of Standards and Technology

 4:10 - 4:40  Break

 4:40 - 5:30  Open Problems and Short Presentations

Wednesday, June 6, 2007

 8:45 -  9:30 Breakfast and Registration

 9:30 - 10:15 Slow Mixing of Glauber Dynamics and Simulated Tempering Algorithms
              Dana Randall, Georgia Institute of Technology

10:15 - 10:45 Break

10:45 - 11:30 A Theoretical Comparison of the Data Augmentation, Marginal
              Augmentation, and PX-DA Algorithms
              Jim Hobert, University of Florida
 
11:40 - 12:25 Implementing Gibbs-type Samplers Using Incompatible Draws, 
              with Applications in High-Energy Astrophysics
              David van Dyk, University of California, Irvine

12:30 -  2:30 Lunch

 2:30 -  3:15 Centers for Random Walks on Trees
              Andrew Beveridge, Carnegie Mellon University
           
 3:25 -  4:10 Fixed-Width Output Analysis for Markov Chain Monte Carlo
              Galin Jones, University of Minnesota

 4:10 - 4:40  Break

 4:40 - 5:30  Open Problems and Short Presentations

 6:00         Banquet

Thursday, June 7, 2007

 8:45 -  9:30 Breakfast and Registration

 9:30 - 10:15 Mixing of Gibbs Sampling on Random Graphs
              Elchanan Mossel, University of California, Berkeley

10:15 - 10:45 Break

10:45 - 11:30 Biopolymer Structure Simulation and Optimization via Fragment Re-growth Monte Carlo
              Jinfeng Zhang, Harvard University

11:40 - 12:25 Exact and Approximate Monte Carlo for Spatial Models
              Murali Haran, Pennsylvania State University

12:30 -  2:30 Lunch

 2:30 -  3:15 Can Extra Updates Delay Mixing?
              Yuval Peres, Microsoft Research and University of California, Berkeley
           
 3:25 -  4:10 Efficiency of Adaptive MCMC
              Yves Atchadé, University of Michigan

 4:10 -  4:40 Break

 4:40 -  5:30 Open Problems and Short Presentations

Previous: Participation

Next: Registration

Workshop Index

DIMACS Homepage