On the rate of convergence of the maximum likelihood estimator in Brownian semimartingale models
Van Zanten, Harry
Bernoulli, Tome 11 (2005) no. 1, p. 643-664 / Harvested from Project Euclid
In this paper we present a unified approach to obtaining rates of convergence for the maximum likelihood estimator (MLE) in Brownian semimartingale models of the form \[ \d X_t = \beta^{n,\theta}_t\,\d t + \sigma^n_t\,\d W_t, \qquad t \le T_n. \] We show that the rate of the MLE is determined by (an appropriate version of) the entropy of the parameter space with respect to the random metric hn, defined by \[ h^2_n(\theta, \psi) = \int_0^{T_n}\left(\frac{\beta^{n,\theta}_s-\beta^{n,\psi}_s}{\sigma^n_s}\right)^2 \,\d s. \] Several known results for the rates in certain popular sub-models of the Brownian semimartingale model are shown to be special cases in our general framework.
Publié le : 2005-08-14
Classification:  continuous semimartingale,  entropy,  exponential inequalities,  maximum likelihood estimation,  rate of convergence
@article{1126126763,
     author = {Van Zanten, Harry},
     title = {On the rate of convergence of the maximum likelihood estimator in Brownian semimartingale models},
     journal = {Bernoulli},
     volume = {11},
     number = {1},
     year = {2005},
     pages = { 643-664},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1126126763}
}
Van Zanten, Harry. On the rate of convergence of the maximum likelihood estimator in Brownian semimartingale models. Bernoulli, Tome 11 (2005) no. 1, pp.  643-664. http://gdmltest.u-ga.fr/item/1126126763/