The likelihood ratio test for general mixture models with or without structural parameter

Jean-Marc Azaïs; Élisabeth Gassiat; Cécile Mercadier

ESAIM: Probability and Statistics (2009)

  • Volume: 13, page 301-327
  • ISSN: 1292-8100

Abstract

top
This paper deals with the likelihood ratio test (LRT) for testing hypotheses on the mixing measure in mixture models with or without structural parameter. The main result gives the asymptotic distribution of the LRT statistics under some conditions that are proved to be almost necessary. A detailed solution is given for two testing problems: the test of a single distribution against any mixture, with application to Gaussian, Poisson and binomial distributions; the test of the number of populations in a finite mixture with or without structural parameter.

How to cite

top

Azaïs, Jean-Marc, Gassiat, Élisabeth, and Mercadier, Cécile. "The likelihood ratio test for general mixture models with or without structural parameter." ESAIM: Probability and Statistics 13 (2009): 301-327. <http://eudml.org/doc/250649>.

@article{Azaïs2009,
abstract = { This paper deals with the likelihood ratio test (LRT) for testing hypotheses on the mixing measure in mixture models with or without structural parameter. The main result gives the asymptotic distribution of the LRT statistics under some conditions that are proved to be almost necessary. A detailed solution is given for two testing problems: the test of a single distribution against any mixture, with application to Gaussian, Poisson and binomial distributions; the test of the number of populations in a finite mixture with or without structural parameter. },
author = {Azaïs, Jean-Marc, Gassiat, Élisabeth, Mercadier, Cécile},
journal = {ESAIM: Probability and Statistics},
keywords = {Likelihood ratio test; mixture models; number of components; local power; contiguity; likelihood ratio test; number of components},
language = {eng},
month = {7},
pages = {301-327},
publisher = {EDP Sciences},
title = {The likelihood ratio test for general mixture models with or without structural parameter},
url = {http://eudml.org/doc/250649},
volume = {13},
year = {2009},
}

TY - JOUR
AU - Azaïs, Jean-Marc
AU - Gassiat, Élisabeth
AU - Mercadier, Cécile
TI - The likelihood ratio test for general mixture models with or without structural parameter
JO - ESAIM: Probability and Statistics
DA - 2009/7//
PB - EDP Sciences
VL - 13
SP - 301
EP - 327
AB - This paper deals with the likelihood ratio test (LRT) for testing hypotheses on the mixing measure in mixture models with or without structural parameter. The main result gives the asymptotic distribution of the LRT statistics under some conditions that are proved to be almost necessary. A detailed solution is given for two testing problems: the test of a single distribution against any mixture, with application to Gaussian, Poisson and binomial distributions; the test of the number of populations in a finite mixture with or without structural parameter.
LA - eng
KW - Likelihood ratio test; mixture models; number of components; local power; contiguity; likelihood ratio test; number of components
UR - http://eudml.org/doc/250649
ER -

References

top
  1. R.J. Adler, An introduction to continuity, extrema and related topics for general Gaussian processes. Inst. Math. Statist. Lect. Notes-Monograph Ser.12 (1990).  
  2. J.-M. Azais, E. Gassiat C. and Mercadier, Asymptotic distribution and power of the likelihood ratio test for mixtures: bounded and unbounded case. Bernoulli12 (2006) 775–799.  
  3. P.J. Bickel, C.A.J. Klaassen, Y. Ritov and J.A. Wellner, Efficient and adaptive estimation for semiparametric models. Johns Hopkins Series in the Mathematical Sciences, Johns Hopkins University Press, Baltimore, MD (1993).  
  4. A. Chambaz, Testing the order of a model. Ann. Statist.34 (2006) 1166–1203.  
  5. A. Chambaz, A. Garivier and E. Gassiat, A mdl approach to hmm with Poisson and Gaussian emissions. Application to order identification. Submitted (2005).  
  6. H. Chen and J. Chen, Large sample distribution of the likelihood ratio test for normal mixtures, Statist. Probab. Lett.2 (2001) 125–133.  
  7. H. Chen and J. Chen, Test for homogeneity in normal mixtures in the presence of a structural parameter. Statist. Sinica13 (2003) 355–365.  
  8. J. Chen and J.D. Kalbfleisch, Modified likelihood ratio test in finite mixture models with a structural parameter. J. Stat. Planning Inf.129 (2005) 93–107.  
  9. H. Chen, J. Chen and J.D. Kalbfleisch, A modified likelihood ratio test for homogeneity in finite mixture models. J. Roy. Statist. Soc. B63 (2001) 19–29.  
  10. H. Chen, J. Chen and J.D. Kalbfleisch, Testing for a finite mixture model with two components. J. Roy. Statist. Soc. B66 (2004) 95–115.  
  11. H. Chernoff and E. Lander, Asymptotic distribution of the likelihood ratio test that a mixture of two binomials is a single binomial. J. Stat. Planning Inf.43 (1995) 19–40.  
  12. T. Chihara, An introduction to orthogonal polynomials. Gordon and Breach, New York (1978).  
  13. G. Ciuperca, Likelihood ratio statistic for exponential mixtures. Ann. Inst. Statist. Math.54 (2002) 585–594.  
  14. D. Dacunha-Castelle and E. Gassiat, Testing in locally conic models, and application to mixture models. ESAIM Probab. Statist.1 (1997) 285–317.  
  15. D. Dacunha-Castelle and E. Gassiat, Testing the order of a model using locally conic parameterization: population mixtures and stationary ARMA processes. Ann. Statist.27 (1999) 1178–1209.  
  16. C. Delmas, On likelihood ratio test in Gaussian mixture models, Sankya65 (2003) 513-531.  
  17. B. Garel, Likelihood Ratio Test for Univariate Gaussian Mixture. J. Statist. Planning Inf.96 (2001) 325–350.  
  18. B. Garel, Asymptotic theory of the likelihood ratio test for the identification of a mixture. J. Statist. Planning Inf.131 (2005) 271–296.  
  19. E. Gassiat, Likelihood ratio inequalities with applications to various mixtures. Ann. Inst. H. Poincaré Probab. Statist.6 (2002) 897–906.  
  20. E. Gassiat and C. Keribin, The likelihood ratio test for the number of components in a mixture with Markov regime, 2000. ESAIM Probab. Stat.4 (2000) 25–52.  
  21. J. Ghosh and P. Sen, On the asymptotic performance of the log likelihood ratio statistic for the mixture model and related results, Proceedings of the Berkeley Conference in Honor of Jerzy Neyman and Jack Kiefer, Vol. II. Wadsworth, Belmont, CA (1985) 789–806.  
  22. P. Hall and M. Stewart, Theoretical analysis of power in a two-component normal mixture model. J. Statist. Planning Inf.134 (2005) 158–179.  
  23. J.A. Hartigan, A failure of likelihood asymptotics for normal mixtures, In Proceedings of the Berkeley conference in honor of Jerzy Neyman and Jack Kiefer (Berkeley, CA, 1983), Vol. II. Wadsworth, Belmont, CA (1985) 807–810.  
  24. J. Henna, Estimation of the number of components of finite mixtures of multivariate distributions. Ann. Inst. Statist. Math.57 (2005) 655–664.  
  25. L.F. James, C.E. Priebe and D.J. Marchette, Consistent Estimation of Mixture Complexity. Ann. Statist.29 (2001) 1281–1296.  
  26. C. Keribin, Consistent estimation of the order of mixture models. Sankhyā Ser. A62 (2000) 49–66.  
  27. M. Lemdani and O. Pons, Likelihood ratio test for genetic linkage. Statis. Probab. Lett.33 (1997) 15–22.  
  28. M. Lemdani and O. Pons, Likelihood ratio in contamination models. Bernoulli5 (1999) 705–719.  
  29. B.G. Lindsay, Mixture models: Theory, geometry, and applications. NSF-CBMS Regional Conf. Ser. Probab. Statist., Vol. 5. Hayward, CA, Institute for Mathematical Statistics (1995).  
  30. X. Liu and Y. Shao, Asymptotics for the likelihood ratio test in two-component normal mixture models. J. Statist. Planning Inf.123 (2004) 61–81.  
  31. X. Liu, C. Pasarica and Y. Shao, Testing homogeneity in gamma mixture models. Scand. J. Statist.30 (2003) 227–239.  
  32. Y. Lo, Likelihood ratio tests of the number of components in a normal mixture with unequal variances. Statis. Probab. Lett.71 (2005) 225–235.  
  33. F. Lord, Estimating the true-score distributions in psychological testing (an empirical bayes estimation problem). Psychometrika34 (1969) 259–299.  
  34. G. McLachlan and D. Peel, Finite mixture models Wiley Series in Probability and Statistics: Applied Probability and Statistics. Wiley-Interscience, New York (2000).  
  35. C. Mercadier (2005), toolbox MATLAB. http://www.math.univ-lyon1.fr/mercadier/MAGP/  
  36. N. Misra, H. Singh and E.J. Harner, Stochastic comparisons of poisson and binomial random varaibles with their mixtures. Statist. Probab. Lett.65 279–290.  
  37. S.A. Murphy and A.W. van der Vaart, Semiparametric likelihood ratio inference. Ann. Statist.25 (1997) 1471–1509.  
  38. Y.S. Quin and B. Smith, Likelihood ratio test for homogeneity in normal mixtures in the presence of a structural parameter. Statist. Sinica143 (2004) 1165–1177.  
  39. Y.S. Quin and B. Smith, The likelihood ratio test for homogeneity in bivariate normal mixtures. J. Multivariate Anal.97 (2006) 474–491.  
  40. D.M. Titterington, A.F.M. Smith and U.E. Makov, Statistical analysis of finite mixture distributions. Wiley Series in Probability and Mathematical Statistics: Applied Probability and Statistics. John Wiley & Sons, Ltd (1985).  
  41. A.W. van der Vaart and J.A. Wellner, Weak convergence and empirical processes, Springer Ser. Statist. Springer-Verlag (1996).  
  42. A.W. van der Vaart, Asymptotic statistics, Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press, Cambridge (1998).  
  43. A.W. van der Vaart, Semiparametric Statistics, Lectures on probability theory and statistics, Saint-Flour, 1999. Lect. Notes Math. 1781 331–457. Springer, Berlin (2002).  
  44. G.R. Wood, Binomial mixtures: geometric estimation of the mixing distribution. Ann. Statist.5 (1999) 1706–1721.  

NotesEmbed ?

top

You must be logged in to post comments.

To embed these notes on your page include the following JavaScript code on your page where you want the notes to appear.

Only the controls for the widget will be shown in your chosen language. Notes will be shown in their authored language.

Tells the widget how many notes to show per page. You can cycle through additional notes using the next and previous controls.

    
                

                
Note: Best practice suggests putting the JavaScript code just before the closing </body> tag.