Abstract
Belief propagation (BP) is a message-passing method for solving probabilistic graphical models. It is very successful in treating disordered models (such as spin glasses) on random graphs. On the other hand, finite-dimensional lattice models have an abundant number of short loops, and the BP method is still far from being satisfactory in treating the complicated loop-induced correlations in these systems. Here we propose a loop-corrected BP method to take into account the effect of short loops in lattice spin models. We demonstrate, through an application to the square-lattice Ising model, that loop-corrected BP improves over the naive BP method significantly. We also implement loop-corrected BP at the coarse-grained region graph level to further boost its performance.
Similar content being viewed by others
References
J. Pearl, Probabilistic Reasoning in Intelligent Systems: Networks of Plausible Inference (Morgan Kaufmann, San Franciso, CA, USA, 1988)
M. Mézard, G. Parisi, M.A. Virasoro, Spin Glass Theory and Beyond (World Scientific, Singapore, 1987)
H.A. Bethe, Proc. R. Soc. London A 150, 552 (1935)
R. Peierls, Proc. Camb. Phil. Soc. 32, 477 (1936)
T.S. Chang, Proc. R. Soc. London A 161, 546 (1937)
M. Mézard, A. Montanari, Information, Physics, and Computation (Oxford University Press, New York, 2009)
J.S. Yedidia, W.T. Freeman, Y. Weiss, IEEE Trans. Inf. Theory 51, 2282 (2005)
A. Pelizzola, J. Phys. A 38, R309 (2005)
T. Rizzo, A. Lage-Castellanos, R. Mulet, F. Ricci-Tersenghi, J. Stat. Phys. 139, 375 (2010)
C. Wang, H.-J. Zhou, J. Phys.: Conf. Ser. 473, 012004 (2013)
A. Lage-Castellanos, R. Mulet, F. Ricci-Tersenghi, Europhys. Lett. 107, 57011 (2014)
R. Kikuchi, Phys. Rev. 81, 988 (1951)
G. An, J. Stat. Phys. 52, 727 (1988)
H.-J. Zhou, C. Wang, J. Stat. Phys. 148, 513 (2012)
S.F. Edwards, P.W. Anderson, J. Phys. F 5, 965 (1975)
H. Nishimori, Statistical Physics of Spin Glasses and Information Processing: An Introduction (Clarendon Press, Oxford, 2001)
A. Montanari, T. Rizzo, J. Stat. Mech.: Theor. Exp. 2005, P10011 (2005)
G. Parisi, F. Slanina, J. Stat. Mech.: Theor. Exp. 2006, L02003 (2006)
M. Chertkov, V.Y. Chernyak, J. Stat. Mech.: Theor. Exp. 2006, P06009 (2006)
J. Mooij, B. Wemmenhove, B. Kappen, T. Rizzo, J. Machine Learning Res.: Workshop Conf. Proc. 2, 331 (2007)
Y. Bulatov, Cycle-corrected belief propagation, www.yaroslavvb.com/papers/bulatov-cycle.pdf (2008)
J.-Q. Xiao, H.-J. Zhou, J. Phys. A 44, 425001 (2011)
S. Ravanbakhsh, C.-N. Yu, R. Greiner, A generalized loop correction method for approximate inference in graphical models, in Proc. 29th Int. Conf. Machine Learning (ICML-12) (Edinburgh, Scotland, 2012), 543
D. Weitz, Counting independent sets up to the inference in graphical models, in Proceedings of the Thirty-Eighth Annual ACM Symposium on Theory of Computing (STOC ’06), (ACM, New York, 2006), 140
E. Ising, Z. Phys. 31, 253 (1925)
K. Huang, Statistical Mechanics, 2nd edn. (John Wiley, New York, 1987)
H.-J. Zhou, Spin Glass and Message Passing (Science Press, Beijing, 2015) (in Chinese)
H.A. Kramers, G.H. Wannier, Phys. Rev. 60, 252 (1941)
H.A. Kramers, G.H. Wannier, Phys. Rev. 60, 263 (1941)
A. Lage-Castellananos, R. Mulet, F. Ricci-Tersenghi, T. Rizzo, Phys. Rev. E 84, 046706 (2011)
I. Morgenstern, K. Binder, Phys. Rev. B 22, 288 (1980)
L. Saul, M. Kardar, Phys. Rev. E 48, R3221 (1993)
J. Houdayer, Eur. Phys. J. B 22, 479 (2001)
T. Jörg, J. Lukic, E. Marinari, O.C. Martin, Phys. Rev. Lett. 96, 237205 (2006)
C.K. Thomas, A.A. Middleton, Phys. Rev. E 80, 046708 (2009)
F.P. Toldin, A. Pelissetto, E. Vicari, Phys. Rev. E 82, 021106 (2010)
C.K. Thomas, D.A. Huse, A.A. Middleton, Phys. Rev. Lett. 107, 047203 (2011)
F.P. Toldin, A. Pelissetto, E. Vicari, Phys. Rev. E 84, 051116 (2011)
R. Kikuchi, S.G. Brush, J. Chem. Phys. 47, 195 (1967)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Zhou, HJ., Zheng, WM. Loop-corrected belief propagation for lattice spin models. Eur. Phys. J. B 88, 336 (2015). https://doi.org/10.1140/epjb/e2015-60485-6
Received:
Revised:
Published:
DOI: https://doi.org/10.1140/epjb/e2015-60485-6