- Access by Xinjiang University
Learning and inference in a nonequilibrium Ising model with hidden nodes
Phys. Rev. E 87, 022127 – Published 20 February, 2013
DOI: https://doi.org/10.1103/PhysRevE.87.022127
Abstract
We study inference and reconstruction of couplings in a partially observed kinetic Ising model. With hidden spins, calculating the likelihood of a sequence of observed spin configurations requires performing a trace over the configurations of the hidden ones. This, as we show, can be represented as a path integral. Using this representation, we demonstrate that systematic approximate inference and learning rules can be derived using dynamical mean-field theory. Although naive mean-field theory leads to an unstable learning rule, taking into account Gaussian corrections allows learning the couplings involving hidden nodes. It also improves learning of the couplings between the observed nodes compared to when hidden nodes are ignored.
Article Text
References (21)
- H. J. Kappen and F. B. Rodriguez, Neur. Comp. 10, 1137 (1998); T. Tanaka, Phys. Rev. E 58, 2302 (1998); Y. Roudi, E. Aurell, and J. Hertz, Front. Comp. Neur. 3, 22 (2009); S. Cocco and R. Monasson, Phys. Rev. Lett. 106, 090601 (2011); F. Ricci-Tersenghi, J. Stat. Mech.: Theory and Exp. (2012) P08015; H. C. Nguyen and J. Berg, ibid. (2012) P03004; Phys. Rev. Lett. 109, 050602 (2012).
- Y. Roudi and J. Hertz, Phys. Rev. Lett. 106, 048702 (2011); M. Mézard and J. Sakellariou, J. Stat. Mech.: Theory Exp. (2011) L07001; H.-L. Zeng, E. Aurell, M. Alava, and H. Mahmoudi, Phys. Rev. E. 83, 041135 (2011); P. Zhang, J. Stat. Phys. 148, 502 (2012).
- D. H. Ackley, G. E. Hinton, and T. J. Sejnowski, Cognitive Science 9, 147 (1985).
- E. Schneidman, M. Berry, R. Segev, and W. Bialek, Nature (London) 440, 1007 (2006).
- J. Shlens, G. Field, J. Gauthier, M. Grivich, D. Petrusca, A. Sher, A. Litke, and E. Chichilnisky, J. Neurosci. 26, 8254 (2006).
- S. Cocco, S. Leibler, and R. Monasson, Proc. Natl. Acad. Sci. USA 106, 14058 (2009).
- T. R. Lezon, J. R. Banavar, M. Cieplak, A. Maritan, and N. Fedoroff, Proc. Natl. Acad. Sci. USA 103 (2006).
- T. Bury, Physica A 392, 1375 (2012).
- C. Bishop, in Learning in Graphical Models, edited by M. I. Jordan (MIT Press, Cambridge, MA, 1999).
- J. Pearl, Causality: Models, Reasoning, and Inference (Cambridge University Press, Cambridge, MA, 2000).
- D. J. Thouless, P. W. Anderson, and R. G. Palmer, Philos. Mag. 35, 593 (1977).
- A. Coolen, in Neuro-Informatics and Neural Modelling, Handbook of Biological Physics, Vol. 4, edited by F. Moss and S. Gielen (North-Holland, Amersterdam, 2001), pp. 619–684.
- M. Opper and O. Winther, in Advanced Mean Field Methods: Theory and Practice, edited by M. Opper and D. Saad (MIT Press, Cambrdige, MA, 2001), pp. 7–21.
- J. Negele and H. Orland, Quantum Many-Particle Systems (Perseus Books, New York, 1998).
- A. Kholodenko, J. Stat. Phys. 58, 357 (1990).
- S. Osindero and G. Hinton, in Advances in Neural Information Processing Systems (MIT Press, Cambridge, MA, 2008), Vol. 20, pp. 1121–1128; G. W. Taylor and G. Hinton, in Proceedings of the 26th Annual International Conference on Machine Learning (Omnipress, Montreal, Quebec, 2009), pp. 1025–1032.
- T. Plefka, J. Phys. A: Math. Gen. 15, 1971 (1981); G. Biroli, ibid. 32, 8365 (1999); Y. Roudi and J. Hertz, J. Stat Mech: Theory Exp. (2011) P03031.
- H. J. Kappen and J. J. Spanjers, Phys. Rev. E 61, 5658 (2000).
- J. Tyrcha and J. Hertz, arXiv:1301.7274.
- C. De Dominicis, Phys. Rev. B 18, 4913 (1978); H. Sompolinsky and A. Zippelius, Phys. Rev. Lett. 47, 359 (1981).
- K. Fischer and J. A. Hertz, Spin Glasses (Cambridge University Press, Cambridge, UK, 1991).