B0443
Title: Adaptive maximum likelihood type estimators for discretely observed small diffusion processes
Authors: Masayuki Uchida - Osaka University (Japan) [presenting]
Abstract: Parameter estimation is considered for multi-dimensional diffusion processes with small dispersion parameters from discrete observations. The joint estimation of both the drift and diffusion parameters of diffusion processes with small dispersion parameters was previously investigated under the general conditions on the sample size and the small dispersion parameter. We propose two kinds of adaptive maximum likelihood type estimators for both the drift and diffusion parameters of diffusion processes with small dispersion parameters from the viewpoint of numerical computation. It is also shown that the proposed estimators of both the drift and diffusion parameters have asymptotic normality under the same general conditions on the sample size and the small dispersion parameter as previously. Furthermore, some examples and simulation results of the proposed drift and diffusion parameters estimators are given.
