The Annals of Statistics

Convex Models, MLS and Misspecification

Valentin Patilea

Full-text: Access has been disabled (more information)

Abstract

We analyze the asymptotic behavior of maximum likelihood estimators (MLE) in convex dominated models when the true distribution generating the independent data does not necessarily belong to the model. Inspired by the Hellinger distance and its properties, we introduce a family of divergences (contrast functions) which allow a unified treatment of well- and misspecified convex models. Convergence and rates of convergence of the MLE with respect to our divergences are obtained from inequalities satisfied by these divergences and results from empirical process theory (uniform laws of large numbers and maximal inequalities). As a particular case we recover existing results for Hellinger convergence of MLE in well-specified convex models. Four examples are considered: mixtures of discrete distributions, monotone densities, decreasing failure rate distributions and a finite-dimensional parametric model.

Article information

Source
Ann. Statist. Volume 29, Number 1 (2001), 94-123.

Dates
First available in Project Euclid: 5 August 2001

Permanent link to this document
http://projecteuclid.org/euclid.aos/996986503

Digital Object Identifier
doi:10.1214/aos/996986503

Mathematical Reviews number (MathSciNet)
MR1833960

Zentralblatt MATH identifier
1029.62020

Subjects
Primary: 62A10
Secondary: 62G20: Asymptotic properties 62F12: Asymptotic properties of estimators

Keywords
convex models maximum likelihood misspecification empirical process

Citation

Patilea, Valentin. Convex Models, MLS and Misspecification. Ann. Statist. 29 (2001), no. 1, 94--123. doi:10.1214/aos/996986503. http://projecteuclid.org/euclid.aos/996986503.


Export citation

References

  • [1] Barlow, R. E., Bartholomew, D. J., Bremner, J. M. and Brunk, H. D. (1972). Statistical Inference under Order Restrictions. Wiley, New York.
  • [2] Barlow, R. E, Marshall, A. and Prochan, F. (1963). Properties of probability distributions with monotone hazard rates. Ann. Math. Statist. 34 375-389.
    Mathematical Reviews (MathSciNet): MR30:1559
    Zentralblatt MATH: 0249.60006
    Digital Object Identifier: doi:10.1214/aoms/1177704147
    Project Euclid: euclid.aoms/1177704147
  • [3] Billingsley, P. (1968). Convergence of Probability Measures. Wiley, New York.
    Mathematical Reviews (MathSciNet): MR38:1718
  • [4] Birg´e, L. (1989). The Grenander estimator: a nonasymptotic approach. Ann. Statist. 17 1532-1549.
    Mathematical Reviews (MathSciNet): MR91d:62035
    Zentralblatt MATH: 0703.62042
    Digital Object Identifier: doi:10.1214/aos/1176347380
    Project Euclid: euclid.aos/1176347380
  • [5] Birg´e, L. and Massart, P. (1993). Rates of convergence for minimum contrast estimators. Probab. Theory Related Fields 97 113-150.
    Mathematical Reviews (MathSciNet): MR94m:62095
    Zentralblatt MATH: 0805.62037
    Digital Object Identifier: doi:10.1007/BF01199316
  • CONVEXITY, MLE AND MISSPECIFICATION 123
  • [6] Dahlhaus, R. and Wefelmeyer, W. (1996). Asymptotically optimal estimation in misspecified time series models. Ann. Statist. 24 952-974.
    Mathematical Reviews (MathSciNet): MR97h:62078
    Zentralblatt MATH: 0865.62063
    Digital Object Identifier: doi:10.1214/aos/1032526951
    Project Euclid: euclid.aos/1032526951
  • [7] Grenander, U. (1956). On the theory of mortality measurement, part II. Scand. Actuar. J. 39 125-153.
    Mathematical Reviews (MathSciNet): MR93415
    Zentralblatt MATH: 0077.33715
  • [8] Huber, P. (1967). The behavior of maximum likelihood estimates under nonstandard conditions. Proc. Fifth Berkeley Symp. Math. Statist. Probab. 1 221-233. Univ. California Press, Berkeley.
    Mathematical Reviews (MathSciNet): MR35:7449
    Zentralblatt MATH: 0212.21504
  • [9] Lambert, D. and Tierney, L. (1984). Asymptotic properties of the maximum likelihood estimates in mixed Poisson model. Ann. Statist. 12 1388-1399.
    Mathematical Reviews (MathSciNet): MR86f:62037
    Zentralblatt MATH: 0562.62038
    Digital Object Identifier: doi:10.1214/aos/1176346799
    Project Euclid: euclid.aos/1176346799
  • [10] Lindsay B. (1983). The geometry of mixture likelihood: a general theory. Ann. Statist. 11 86-94.
    Mathematical Reviews (MathSciNet): MR684866
    Zentralblatt MATH: 0512.62005
    Digital Object Identifier: doi:10.1214/aos/1176346059
    Project Euclid: euclid.aos/1176346059
  • [11] Koltchinskii, V. I. (1997). M-estimation, convexity and quantiles. Ann. Statist. 25 435-477.
    Mathematical Reviews (MathSciNet): MR98m:62181
    Zentralblatt MATH: 0878.62037
    Digital Object Identifier: doi:10.1214/aos/1031833659
    Project Euclid: euclid.aos/1031833659
  • [12] L ¨uxmann-Ellinghaus, U. (1987). On the identifiability of mixture of infinitely divisible power series distributions. Statist. Probab. Lett. 5 375-379.
    Mathematical Reviews (MathSciNet): MR88k:60031
    Zentralblatt MATH: 0631.60013
  • [13] Marshall, A. W. (1970). Discussion of Barlow and van Zwet's paper "Asymptotic properties of isotonic estimators for the generalized failure rate function: strong consistency." In Nonparametric Techniques in Statistical Inference (M. L. Puri, ed.) 174-176. Cambridge Univ. Press.
    Mathematical Reviews (MathSciNet): MR275607
  • [14] Marshall, A. W. and Proschan, F. (1965). Maximum likelihood estimation for distribution with monotone failure rate. Ann. Math. Statist. 36 69-77.
    Mathematical Reviews (MathSciNet): MR30:674
    Zentralblatt MATH: 0128.38506
    Digital Object Identifier: doi:10.1214/aoms/1177700271
    Project Euclid: euclid.aoms/1177700271
  • [15] Noack, A. (1950). A class of random variables with discrete distribution Ann. Math. Statist. 21 127-132.
    Mathematical Reviews (MathSciNet): MR11,445g
    Zentralblatt MATH: 0036.08601
    Digital Object Identifier: doi:10.1214/aoms/1177729894
    Project Euclid: euclid.aoms/1177729894
  • [16] Milhaud, X. and Mounime, S. (1996). A modified maximum likelihood estimator for infinite mixtures. Preprint, Univ. P. Sabatier, Toulouse.
  • [17] Patilea, V. (1997). Convex models, NPLME and misspecification, Ph.D dissertation, part I. Institute of Statistics, Univ. catholique de Louvain.
  • [18] Pfanzagl J. (1969). On the measurability and consistency of minimum contrast estimates. Metrika 249-272.
  • [19] Pfanzagl J. (1988). Consistency of maximum likelihood estimators for certain nonparametric families, in particular: Mixtures. J. Statist. Plann. Inference 19 137-158.
    Mathematical Reviews (MathSciNet): MR944202
    Zentralblatt MATH: 0656.62044
    Digital Object Identifier: doi:10.1016/0378-3758(88)90069-9
  • [20] Pfanzagl, J. (1990). Large deviations probabilities for certain nonparametric maximum likelihood estimators. Ann. Statist. 18 1868-1877.
    Mathematical Reviews (MathSciNet): MR92d:62028
    Zentralblatt MATH: 0721.62048
    Digital Object Identifier: doi:10.1214/aos/1176347884
    Project Euclid: euclid.aos/1176347884
  • [21] Pollard, D. (1984). Convergence of Stochastic Processes. Springer, New York.
    Mathematical Reviews (MathSciNet): MR86i:60074
    Zentralblatt MATH: 0544.60045
  • [22] Rolin, J.-M. (1992). Personal notes.
  • [23] Roberts, W. and Varberg, D. (1973). Convex Functions. Academic Press, New York.
    Mathematical Reviews (MathSciNet): MR442824
  • [24] Simar, L. (1976). Maximum likelihood estimation of a compound Poisson process. Ann. Statist. 4 1200-1209.
    Mathematical Reviews (MathSciNet): MR55:1603
    Zentralblatt MATH: 0362.62095
    Digital Object Identifier: doi:10.1214/aos/1176343651
    Project Euclid: euclid.aos/1176343651
  • [25] van de Geer, S. (1993). Hellinger consistency of certain nonparametric maximum likelihood estimators. Ann. Statist. 21 14-44.
    Mathematical Reviews (MathSciNet): MR94c:62062
    Zentralblatt MATH: 0779.62033
    Digital Object Identifier: doi:10.1214/aos/1176349013
    Project Euclid: euclid.aos/1176349013
  • [26] van de Geer, S. (2000). Empirical Processes in M-estimation. Cambridge Univ. Press.
  • [27] van der Vaart, A. and Wellner, J. (1996). Weak Convergence and Empirical Process. Springer, New York.
    Mathematical Reviews (MathSciNet): MR97g:60035
    Zentralblatt MATH: 0862.60002
  • [28] Wang, J.-L. (1985). Strong consistency of approximate maximum likelihood estimators with applications to nonparametrics. Ann. Statist. 13 932-946.
    Mathematical Reviews (MathSciNet): MR803749
    Zentralblatt MATH: 0598.62034
    Digital Object Identifier: doi:10.1214/aos/1176349647
    Project Euclid: euclid.aos/1176349647
  • [29] Wong, W. H. and Shen, X. (1995). Probabilities inequalities for likelihood ratios and convergence rates of sieve MLE's. Ann. Statist. 23 339-362.
    Mathematical Reviews (MathSciNet): MR1332570
    Zentralblatt MATH: 0829.62002
    Digital Object Identifier: doi:10.1214/aos/1176324524
    Project Euclid: euclid.aos/1176324524

The Institute of Mathematical Statistics

The Annals of Statistics

New content alerts

Email RSS ToC RSS Article
  • You have access to this content.
  • You have partial access to this content.
  • You do not have access to this content.
Turn Off MathJax

What is MathJax?

More like this

+ See more