Statistics and Computing

, 19:283 | Cite as

Pairwise likelihood estimation for multivariate mixed Poisson models generated by Gamma intensities

  • Florent Chatelain
  • Sophie Lambert-LacroixEmail author
  • Jean-Yves Tourneret


Estimating the parameters of multivariate mixed Poisson models is an important problem in image processing applications, especially for active imaging or astronomy. The classical maximum likelihood approach cannot be used for these models since the corresponding masses cannot be expressed in a simple closed form. This paper studies a maximum pairwise likelihood approach to estimate the parameters of multivariate mixed Poisson models when the mixing distribution is a multivariate Gamma distribution. The consistency and asymptotic normality of this estimator are derived. Simulations conducted on synthetic data illustrate these results and show that the proposed estimator outperforms classical estimators based on the method of moments. An application to change detection in low-flux images is also investigated.


Pairwise likelihood estimation Multivariate mixed Poisson models Multivariate Gamma distributions Negative multinomial distributions 


  1. Baddeley, A., Turner, R.: Practical maximum pseudolikelihood for spatial point patterns (with discussion). Austral. New Zealand J. Stat. 42(3), 283–322 (2000) zbMATHCrossRefMathSciNetGoogle Scholar
  2. Bernardoff, P.: Which negative multinomial distributions are infinitely divisible? Bernoulli 9(5), 877–893 (2003) zbMATHCrossRefMathSciNetGoogle Scholar
  3. Bernardoff, P.: Which multivariate Gamma distributions are infinitely divisible? Bernoulli 12(1), 169–189 (2006) zbMATHMathSciNetGoogle Scholar
  4. Besag, J.: Spatial interaction and the statistical analysis of lattice systems (with discussion). J. R. Stat. Soc. Ser. B 36(2), 192–236 (1974) zbMATHMathSciNetGoogle Scholar
  5. Besag, J.: Some methods of statistical analysis for spatial data. Bull. Int. Stat. Inst. 47, 77–92 (1977) MathSciNetGoogle Scholar
  6. Besag, J., Milne, R., Zachary, S.: Point process limits of lattice processes. J. Appl. Probab. 19, 210–216 (1982) zbMATHCrossRefMathSciNetGoogle Scholar
  7. Le Cessie, S., Van Houwelingen, J.C.: Logistic regression for correlated binary data. Appl. Stat. 43(1), 95–108 (1994) zbMATHCrossRefGoogle Scholar
  8. Chatelain, F., Ferrari, A., Tourneret, J.-Y.: Parameter estimation for multivariate mixed Poisson distributions. In: Proc. IEEE ICASSP-06, vol. IV, pp. 17–20, Toulouse, France, May 2006 Google Scholar
  9. Chatelain, F., Tourneret, J.-Y., Ferrari, A., Inglada, J.: Bivariate Gamma distributions for image registration and change detection. IEEE Trans. Image Process. 16(7), 1796–1806 (2007). Available at CrossRefMathSciNetGoogle Scholar
  10. Chatelain, F., Tourneret, J.-Y., Inglada, J.: Change detection in multisensor SAR images using multivariate Gamma distributions. IEEE Trans. Image Process. 17(3), 249–258 (2008). Available at CrossRefMathSciNetGoogle Scholar
  11. Cox, D.R., Reid, N.: A note on pseudolikelihood constructed from marginal densities. Biometrika 91(3), 729–737 (2004) zbMATHCrossRefMathSciNetGoogle Scholar
  12. Dacunha-Castelle, D., Duflo, M.: Probabilités et Statistiques, Tome 2. Masson, Paris (1984) zbMATHGoogle Scholar
  13. Fearnhead, P.: Consistency of estimators of the population-scaled recombination rate. Theor. Population Biology 64(13), 67–79 (2003) zbMATHCrossRefGoogle Scholar
  14. Ferrari, A., Letac, G., Tourneret, J.-Y.: Multivariate mixed Poisson distributions. In: Hlawatsch, F., Matz, G., Rupp, M., Wistawel, B. (eds.) EUSIPCO-04,, Vienna, Austria, September 2004, pp. 1067–1070. Elsevier, Amsterdam (2004) Google Scholar
  15. Ferrari, A., Letac, G., Tourneret, J.-Y.: Exponential families of mixed Poisson distributions. J. Multivar. Anal. 98(6), 1283–1292 (2007) zbMATHCrossRefMathSciNetGoogle Scholar
  16. Goodman, J.: Statistical Optics. Wiley, New York (1985) Google Scholar
  17. Goudail, F., Roux, N., Réfrégier, P.: Performance parameters for detection in low-flux coherent images. Opt. Lett. 28(2), 81–83 (2003) CrossRefGoogle Scholar
  18. Grandell, J.: Mixed Poisson Processes. Chapman & Hall, Boca Raton (1997) zbMATHGoogle Scholar
  19. Guan, Y.: A composite likelihood approach in fitting spatial point process models. J. Am. Stat. Assoc. 101, 1502–1512 (2006) zbMATHCrossRefGoogle Scholar
  20. Hall, P., Jing, B.: On sample reuse methods for dependent data. J. R. Stat. Soc. Ser. B 58(4), 727–737 (1996) zbMATHMathSciNetGoogle Scholar
  21. Heagerty, P.J., Lele, S.R.: A composite likelihood approach to binary spatial data. J. Am. Stat. Assoc. 93(443), 1099–1111 (1998) zbMATHCrossRefMathSciNetGoogle Scholar
  22. Henderson, R., Shimakura, S.: A serially correlated gamma frailty model for longitudinal count data. Biometrika 90(2), 355–366 (2003). doi: 10.1093/biomet/90.2.355 zbMATHCrossRefMathSciNetGoogle Scholar
  23. Huang, F., Ogata, Y.: Generalized pseudo-likelihood estimates for Markov random fields on lattice. Ann. Inst. Stat. Math. 54(1), 1–18 (2002) zbMATHCrossRefMathSciNetGoogle Scholar
  24. Jensen, J., Kunsch, R.: On asymptotic normality of pseudo likelihood estimates for pairwise interaction processes. Ann. Inst. Stat. Math. 46, 475–486 (1994) zbMATHMathSciNetGoogle Scholar
  25. Jensen, J., Moller, J.: Pseudolikelihood for exponential family models of spatial point processes. Ann. Appl. Probab. 1(3), 445–461 (1991) zbMATHCrossRefMathSciNetGoogle Scholar
  26. Johnson, N.L., Kotz, S., Kemp, A.W.: Univariate Discrete Distributions. Wiley, New York (1993) Google Scholar
  27. Karlis, D., Xekalaki, E.: Mixed Poisson distributions. Int. Stat. Rev. 73(1), 35–58 (2005) zbMATHGoogle Scholar
  28. Kuk, A.Y.C.: A hybrid pairwise likelihood method. Biometrika 94(4), 939–952 (2007) zbMATHCrossRefMathSciNetGoogle Scholar
  29. Kuk, A.Y.C., Nott, D.J.: A pairwise likelihood approach to analyzing correlated binary data. Stat. Probab. Lett. 47(4), 329–335 (2000) zbMATHCrossRefGoogle Scholar
  30. Leemis, L.M., Trivedi, K.S.: A comparison of approximate interval estimators for the Bernoulli parameter. Am. Stat. 50(1), 63–68 (1996) CrossRefMathSciNetGoogle Scholar
  31. Lindsay, B.G.: Composite likelihood methods. Contemp. Math. 50, 221–239 (1988) Google Scholar
  32. Mateu, J., Montes, F.: Pseudo-likelihood inference for Gibbs processes with exponential families through generalized linear models. Stat. Inference Stoch. Process. 4(2), 125–154 (2001) zbMATHCrossRefMathSciNetGoogle Scholar
  33. Molenberghs, G., Verbeke, G.: Models for Discrete Longitudinal Data, 2nd edn. Springer, Berlin (2006) Google Scholar
  34. Muirhead, R.J.: Aspects of Multivariate Statistical Theory. Wiley-Interscience, New York (1982) zbMATHCrossRefGoogle Scholar
  35. Nott, D.J., Rydén, T.: Pairwise likelihood methods for inference in image models. Biometrika 86(3), 661–676 (1999) zbMATHCrossRefMathSciNetGoogle Scholar
  36. Oliver, C., Quegan, S.: Understanding Synthetic Aperture Radar Images. Artech House, Norwood (1998) Google Scholar
  37. Radke, R.J., Andra, S., Al-Kofahi, O., Roysam, B.: Image change detection algorithms: a systematic survey. IEEE Trans. Image Process. 14(3), 294–307 (2005) CrossRefMathSciNetGoogle Scholar
  38. van der Vaart, A.W.: Asymptotic Statistics. Cambridge University Press, Cambridge (1998) zbMATHGoogle Scholar
  39. Van Trees, H.L.: Detection, Estimation, and Modulation Theory: Part I. Wiley, New York (1968) zbMATHGoogle Scholar
  40. Varin, C.: On composite marginal likelihoods. Adv. Stat. Anal. 92(1), 1–28 (2008) CrossRefMathSciNetzbMATHGoogle Scholar
  41. Varin, C., Vidoni, P.: Pairwise likelihood inference for ordinal categorical time series. Comput. Stat. Data Anal. 51(4), 2365–2373 (2006). ISSN 0167-9473. doi: zbMATHCrossRefMathSciNetGoogle Scholar
  42. Zhao, Y., Joe, H.: Composite likelihood estimation in multivariate data analysis. Canad. J. Stat. 33(3), 335–356 (2005) zbMATHCrossRefMathSciNetGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  • Florent Chatelain
    • 1
  • Sophie Lambert-Lacroix
    • 2
    Email author
  • Jean-Yves Tourneret
    • 1
  1. 1.IRIT/ENSEEIHT/TésaToulouse Cedex 7France
  2. 2.Laboratoire Jean KuntzmannUniversité de Grenoble et CNRSGrenoble Cedex 9France

Personalised recommendations