SPEAKER: |
|
TITLE: |
"Perfect sampling: not just for Markov chains?" |
DAY: |
Wednesday, November 17, 1999 |
TIME: |
3:30 p.m. [Coffee & cookies in BSB-202 at 3:00 p.m.] |
PLACE: |
BSB-108 |
Propp and Wilson's (1996) coupling from the past (CFTP) algorithm generates a sample from the limiting distribution of a Markov chain. In this talk I argue that the underlying idea of CFTP is more widely applicable, and will demonstrate attempts to apply it to two problems: approximation of limits and simulation of stochastic differential equations.
Duncan Murdoch earned his BMath from University of Waterloo. His MSc and PhD degrees were obtained at Carleton University. He did his PhD part time while working as a statistical consultant at Health and Welfare Canada in Ottawa. Subsequently Dr Murdoch worked in a research position at University of Waterloo, at Queen's University, and presently at The University of Western Ontario. Dr. Murdoch is the Program Organizer for the 2000 Annual Meeting of the Statistical Society of Canada to be held in Ottawa next year. He has pretty wide interests including biostatistics, statistical computing, statistical graphics and directional data. |
|
The references below, which Dr Murdoch has suggested as useful background for his talk, have been placed on reserve at Thode Library (STATS 770: Statistics Seminar):
[1] Propp, J.G. & Wilson, D.B. (1996) "Exact Sampling With Coupled Markov Chains and Applications to Statistical Mechanics," RANDOM STRUCTURES AND ALGORITHMS 9 (1&2), pp. 223--252.
[2] Propp, J.G. & Wilson, D.B. (1998) "Coupling From the Past: A User's Guide," in D. Aldous and J. Propp, editors, MICROSURVEYS IN DISCRETE PROBABILITY, Vol. 41 of DIMACS Series in Discrete Mathematics and Theoretical Computer Science (American Mathematical Society), pp. 181--192.