Summary
The paper is concerned with the exact simulation of an unobserved true point process conditional on a noisy observation. We use dominated coupling from the past (CFTP) on an augmented state space to produce perfect samples of the target marked point process. An optimized coupling of the target chains makes the algorithm considerable faster than with the standard coupling used in dominated CFTP for point processes. The perfect simulations are used for inference and the results are compared to an ordinary Metropolis-Hastings sampler.
Similar content being viewed by others
References
Cai, Y. & Kendall, W. S. (2002), ‘Perfect simulation for correlated Poisson random variables conditioned to be positive’,Statistics and Computing 12, 229–243.
Dralle, K. & Rudemo, M. (1996), ‘Stem number estimation by kernel smoothing of aerial photos’,Canad. J. Forest Res. 26, 1228–1236.
Dralle, K. & Rudemo, M. (1997), ‘Automatic estimation of individual tree positions from aerial photos’,Canad. J. Forest Res. 27, 1728–1736.
Fill, J. A.(1998), ‘An interruptible algorithm for perfect sampling via Markov chainsrs,Ann. Appl. Probab.8, 131–162.
Geyer, C. J. & Møller, J. (1994), ‘Simulation procedures and likelihood inference for spatial point processes’,Scand. J. of Statist.21(4), 359–373.
Kendall, W. S. & Møller, J. (2001), ‘Perfect simulation using dominating processes on ordered spaces, with application to locally stable point processes’,Adv.in Appl. Probab.32, 844–865.
Kendall, W. & Thönnes, E. (1999), ‘Perfect simulation in stochastic geometry’,Pattern Recognition 32, 1569–1586.
Larsen, M. & Rudemo, M. (1998), ‘Optimizing templates for finding trees in aerial photographs’,Pattern Recognition Letters 19, 1153–1162.
Lund, J., Penttinen, A. & Rudemo, M. (1999), Bayesian analysis of spatial point patters from noisy observations, Preprint, Department of Mathematics and Physics, The Royal Veterinary and Agricultural University.
Lund, J. & Rudemo, M. (2000), ‘Models for point processes observed with noise’,Biometrika 87(2), 235–249.
Møller, J. (1999), ‘Perfect simulation of conditionally specified models’,J. Roy. Statist. Soc. Ser. B 61(1), 251–264.
Møller, J. (2001), A review on perfect simulation in stochastic geometry,in C. C. H. I. V. Basawa & R. L. Taylor, eds, ‘Selected Proceedings of the Symposium on Inference for Stochastic Processes’, Vol. 37 ofIMS Lecture Notes & Monographs, pp. 333–355.
Møller, J. & Nicholls, G. K. (1999), Perfect simulation for sample-based inference, Research report R-99–2011, Aalborg University, Department of Mathematics. To appear in Statistics and Computing (conditionally accepted).
Møller, J. (1999), Markov Chain Monte Carlo and Spatial Point Processes,in O. Barndorff-Nielsen, W. Kendall & M. van Lieshout, eds, ‘Stochastic Geometry: likelihood and computation’, Chapman & Hall/CRC, Boca Raton, pp. 141–172.
Murdoch, D. J. & Green, P. J. (1998), ‘Exact sampling from a continuous state space’, Scand. J. of Statist.25, 483–502.
Preston, C. (1975), ‘Spatial birth-and-death processes’,Bull. Internal. Statist. Inst. 46, 371–391.
Propp, J. & Wilson, D. (1996), ‘Exact sampling with coupled Markov chains and applications to statistical mechanics’,Random Structures and Algorithms 9, 223–252.
Strauss, D. (1975), ‘A model for clustering’,Biometrika 63, 467–475.
Acknowledgements
Elke Thönnes thanks for the support by the EU TMR network ERB-FMRXCT96- 0096 and the Stochastic Centre at Chalmers University. Both authors acknowledge the support by the ESF program “Highly structured stochastic systems”. We are grateful to Ib Skovgaard, Wilfrid Kendall and Jesper Møller for inspiring discussions. We also thank Mats Rudemo for helpful comments, and the Danish Forest and Landscape Research Institute for access to the data.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Lund, J., Thönnes, E. Perfect simulation and inference for point processes given noisy observations. Computational Statistics 19, 317–336 (2004). https://doi.org/10.1007/BF02892063
Published:
Issue Date:
DOI: https://doi.org/10.1007/BF02892063