Biometrika Advance Access originally published online on August 5, 2007
Biometrika 2007 94(3):661-672; doi:10.1093/biomet/asm052
Copyright © 2007 Biometrika Trust
Articles |
Recursive computing and simulation-free inference for general factorizable models
Department of Statistics, University of Glasgow, Glasgow G12 8QW, U.K.
Department of Mathematical Sciences, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
nial{at}stats.gla.ac.uk
havard.rue{at}math.ntnu.no
Received for publication 1 August 2005. Revision received 1 February 2007.
We illustrate how the recursive algorithm of Reeves & Pettitt (2004) for general factorizable models can be extended to allow exact sampling, maximization of distributions and computation of marginal distributions. All of the methods we describe apply to discrete-valued Markov random fields with nearest neighbour integrations defined on regular lattices; in particular we illustrate that exact inference can be performed for hidden autologistic models defined on moderately sized lattices. In this context we offer an extension of this methodology which allows approximate inference to be carried out for larger lattices without resorting to simulation techniques such as Markov chain Monte Carlo. In particular our work offers the basis for an automatic inference machine for such models.
Key Words: Autologistic distribution • Exact sampling • Hidden Markov random field • Ising model • Normalizing constant
References
-
Bartolucci F., Besag J. E. A recursive algorithm for Markov random fields. Biometrika (2002) 89:724–30.
Besag J. E. Spatial interaction and the statistical analysis of lattice systems (with Discussion). J. R. Statist. Soc. (1974) B 36:192–236.
Besag J. E. A candidate's formula: A curious result in Bayesian prediction. Biometrika (1989) 76:183.
Besag J. E., York J. C., Mollié A. Bayesian image restoration, with two applications in spatial statistics (with Discussion). Ann. Inst. Statist. Math. (1991) 43:1–59.[CrossRef]
Bozdech Z., Llinás M., Pulliam B. L., Wong E. D., Zhu J., DeRisi J. L. The transcriptome of the intraerythrocytic developmental cycle of plasmodium falciparum. PLoS Biol. (2003) 1:85–100.[ISI]
Buck C. E., Cavanagh W. G., Litton C. D. Computer Applications and Quantitative Methods in Archaeology—Rahtz S., ed. (1988) Oxford: British Archaeological Reports. 151–60. International Series 446. The spatial analysis of site phosphate data.
Gelman A., Meng X.-L. Simulating normalizing constants: from importance sampling to bridge sampling to path sampling. Statist. Sci. (1998) 13:163–85.[CrossRef]
Geyer C. J., Thompson E. A. Constrained Monte Carlo maximum likelihood for dependent data (with Discussion). J. R. Statist. Soc. (1992) B 54:657–99.
Gu M. G., Zhu H.-T. Maximum likelihood estimation for spatial models by Markov chain monte Carlo stochastic approximation. J. R. Statist. Soc. (2001) B 63:339–55.[CrossRef]
Huang F., Ogata Y. Improvements of the maximum pseudo-likelihood estimators in various spatial statistical models. J. Comp. Graph. Statist. (1999) 8:510–30.[CrossRef]
Reeves R., Pettitt A. N. Efficient recursions for general factorisable models. Biometrika (2004) 91:751–7.
Robert C. P., Rydén T., Titterington D. M. Bayesian inference in hidden Markov models through the reversible jump Markov chain Monte Carlo method. J. R. Statist. Soc. (2000) B 62:57–75.[CrossRef]
Rue H. New loss functions in Bayesian imaging. J. Am. Statist. Assoc. (1995) 90:900–08.[CrossRef][ISI]
Rue H., Steinsland I., Erland S. Approximating hidden Gaussian Markov random fields. J. R. Statist. Soc. (2004) B 66:877–92.[CrossRef]
Scott S. L. Bayesian methods for hidden Markov models: Recursive computing in the 21st century. J. Am. Statist. Assoc. (2002) 97:337–51.[CrossRef][ISI]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||