Diffusions with measurement errors. I. Local Asymptotic Normality

Arnaud Gloter; Jean Jacod

ESAIM: Probability and Statistics (2010)

  • Volume: 5, page 225-242
  • ISSN: 1292-8100

Abstract

top
We consider a diffusion process X which is observed at times i/n for i = 0,1,...,n, each observation being subject to a measurement error. All errors are independent and centered Gaussian with known variance pn. There is an unknown parameter within the diffusion coefficient, to be estimated. In this first paper the case when X is indeed a Gaussian martingale is examined: we can prove that the LAN property holds under quite weak smoothness assumptions, with an explicit limiting Fisher information. What is perhaps the most interesting is the rate at which this convergence takes place: it is 1 / n (as when there is no measurement error) when pn goes fast enough to 0, namely npn is bounded. Otherwise, and provided the sequence pn itself is bounded, the rate is (pn / n)1/4. In particular if pn = p does not depend on n, we get a rate n-1/4.

How to cite

top

Gloter, Arnaud, and Jacod, Jean. "Diffusions with measurement errors. I. Local Asymptotic Normality." ESAIM: Probability and Statistics 5 (2010): 225-242. <http://eudml.org/doc/116585>.

@article{Gloter2010,
abstract = { We consider a diffusion process X which is observed at times i/n for i = 0,1,...,n, each observation being subject to a measurement error. All errors are independent and centered Gaussian with known variance pn. There is an unknown parameter within the diffusion coefficient, to be estimated. In this first paper the case when X is indeed a Gaussian martingale is examined: we can prove that the LAN property holds under quite weak smoothness assumptions, with an explicit limiting Fisher information. What is perhaps the most interesting is the rate at which this convergence takes place: it is $1/\sqrt\{n\}$ (as when there is no measurement error) when pn goes fast enough to 0, namely npn is bounded. Otherwise, and provided the sequence pn itself is bounded, the rate is (pn / n)1/4. In particular if pn = p does not depend on n, we get a rate n-1/4. },
author = {Gloter, Arnaud, Jacod, Jean},
journal = {ESAIM: Probability and Statistics},
keywords = {Statistics of diffusions; measurement errors; LAN property.},
language = {eng},
month = {3},
pages = {225-242},
publisher = {EDP Sciences},
title = {Diffusions with measurement errors. I. Local Asymptotic Normality},
url = {http://eudml.org/doc/116585},
volume = {5},
year = {2010},
}

TY - JOUR
AU - Gloter, Arnaud
AU - Jacod, Jean
TI - Diffusions with measurement errors. I. Local Asymptotic Normality
JO - ESAIM: Probability and Statistics
DA - 2010/3//
PB - EDP Sciences
VL - 5
SP - 225
EP - 242
AB - We consider a diffusion process X which is observed at times i/n for i = 0,1,...,n, each observation being subject to a measurement error. All errors are independent and centered Gaussian with known variance pn. There is an unknown parameter within the diffusion coefficient, to be estimated. In this first paper the case when X is indeed a Gaussian martingale is examined: we can prove that the LAN property holds under quite weak smoothness assumptions, with an explicit limiting Fisher information. What is perhaps the most interesting is the rate at which this convergence takes place: it is $1/\sqrt{n}$ (as when there is no measurement error) when pn goes fast enough to 0, namely npn is bounded. Otherwise, and provided the sequence pn itself is bounded, the rate is (pn / n)1/4. In particular if pn = p does not depend on n, we get a rate n-1/4.
LA - eng
KW - Statistics of diffusions; measurement errors; LAN property.
UR - http://eudml.org/doc/116585
ER -

References

top
  1. P. Bickel and Y. Ritov, Inference in hidden Markov models. I. Local asymptotic normality in the stationary case. Bernoulli2 (1996) 199-228.  
  2. P. Bickel and Y. Ritov, Asymptotic normality for the maximum likelihood estimator for general hidden Markov models. Ann. Statist.26 (1998) 1614-1635.  
  3. G. Dohnal, On estimating the diffusion coefficient. J. Appl. Probab.24 (1987) 105-114.  
  4. V. Genon-Catalot and J. Jacod, On the estimation of the diffusion coefficient for multidimensional diffusion processes. Ann. Inst. H. Poincaré Probab. Statist.29 (1993) 119-153.  
  5. V. Genon-Catalot and J. Jacod, Estimation of the diffusion coefficient for diffusion processes: random sampling. Scand. J. Statist.21 (1994) 193-221.  
  6. A. Gloter and J. Jacod, Diffusion with measurement error. II. Optimal estimators (2000).  
  7. E. Gobet, LAMN property for elliptic diffusions (2000).  
  8. J. Jacod and A. Shiryaev, Limit Theorems for Stochastic Processes. Springer-Verlag, Berlin (1987).  
  9. J.L. Jensen and N. Petersen, Asymptotic normality of the Maximum likelihood estimator in state space models. Ann. Statist.27 (1999) 514-535.  
  10. B. Leroux, Maximum likelihood estimation for hidden Markov models. Stochastic Process. Appl.40 (1992) 127-143.  
  11. L. LeCam and G.L. Yang, Asymptotics in Statistics. Springer-Verlag, Berlin (1990).  
  12. M.B. Malyutov and O. Bayborodin, Fitting diffusion and trend in noise via Mercer expansion, in Proc. 7th Int. Conf. on Analytical and Stochastic Modeling Techniques. Hamburg (2000).  
  13. T. Ryden, Consistent and asymptotically normal estimators for hidden Markov models. Ann. Statist.22 (1994) 1884-1895.  

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.