Mixing times for the Swapping Algorithm on the Blume-Emery-Griffiths Model
Probability
2012-06-20 v1
Abstract
We analyze the so called Swapping Algorithm, a parallel version of the well-known Metropolis-Hastings algorithm, on the mean-field version of the Blume-Emery-Griffiths model in statistical mechanics. This model has two parameters and depending on their choice, the model exhibits either a first, or a second order phase transition. In agreement with a conjecture by Bhatnagar and Randall we find that the Swapping Algorithm mixes rapidly in presence of a second order phase transition, while becoming slow when the phase transition is first order.
Cite
@article{arxiv.1206.4162,
title = {Mixing times for the Swapping Algorithm on the Blume-Emery-Griffiths Model},
author = {M. Ebbers and H. Knöpfel and M. Löwe and F. Vermet},
journal= {arXiv preprint arXiv:1206.4162},
year = {2012}
}
Comments
35 pages, to be published in Random Structures and Algorithms