Avviso di seminario

[Paola Vicard <vicard@uniroma3.it>]

************** AVVISO DI SEMINARIO ***************

Mercoledi' 5 novembre ore 15,00 Aula 9

Il Prof. J. Møller, Department of Mathematical Sciences, Aalborg University

terrà presso il Dipartimento di Economia, Università Roma Tre, Via Ostiense 
139, un seminario dal titolo

AN INTRODUCTION TO PERFECT SIMULATION FOR SPATIAL PROCESSES


Riassunto

Markov chain Monte Carlo algorithms can fail to converge in
practice. A recent topic, which has drawn great attention after the seminal 
work
of Propp and Wilson (1996), is ``exact'' or ``perfect'' simulation. A
simulation algorithm is said to be perfect when one is assured
that equilibrium has been attained when the algorithm finish; here the
algorithm running time is allowed to be random. This is
obviously appealing and useful for several reasons: One needs not to
worry about whether one has used an appropriate burn in before
sampling; i.i.d. sampling is possible so that e.g. asymptotic
variances of Monte Carlo estimates can be straightforwardly
calculated; and one can compare other ``non-perfect''
algorithms with perfect simulations. Propp and Wilson's ideas have now
been extended in many ways, and one area, where perfect simulation
techniques seem particular useful, is spatial statistics.
In the talk I'll review these recent developments on perfect
simulation for spatial processes based on the following and other 
contributions:

J.A. Fill (1998). An interruptible algorithm for perfect sampling
via Markov chains. Annals of Applied Probability, 8, 131 - 162.

O. Häggström, M.N.M. van Lieshout and J. Møller. Characterization results and
Markov chain Monte Carlo algorithms including exact simulation for some 
spatial
point processes. Bernoulli, 5, 641 - 659.

J. Møller. Perfect simulation of conditionally specified models. Journal of 
Royal
Statistical Society Ser. B, 61, 251 - 264.

J. Møller and K. Schladitz. Extensions of Fill's algorithm for perfect 
simulation.
Journal of the Royal Statistical Society Ser. B, 61, 955 - 969.

J.G. Propp and D.B. Wilson (1996). Exact sampling with coupled
Markov chains and applications to statistical mechanics. Random
Structures and Algorithms,  9, 223-252.

____________________________________________________________________
Tutti gli interessati sono invitati a partecipare.

__________________________
Paola Vicard
Dipartimento di Economia
Universita' Roma Tre
Via Ostiense 139, 00154 Roma
Tel: 06 57374219
Fax: 06 57374093
e-mail: vicard@uniroma3.it
___________________________