|
- 2017
随机微分方程的几种参数估计方法
|
Abstract:
摘要 提出3种基于离散观测数据的随机微分方程参数估计的方法。第1种方法应用于线性随机微分方程。推导出这类方程的真解的相关运算服从的分布,使观测数据的运算也服从此分布,由此来估计漂移系数与扩散系数中的未知参数。第2种方法用于It?型随机微分方程。推导出Euler-Maruyama格式的数值解的相关运算服从的分布,使观测数据的运算服从此分布,由此来估计参数。第3种方法用于Stratonovich型随机微分方程。推导出中点格式的数值解的相关运算服从的分布,使观测数据的运算服从此分布,以此来估计参数。数值实验验证了这3种方法的有效性。数值实验显示,Euler-Maruyama格式参数估计的误差约为O(h0.5)阶,中点格式参数估计的误差约为O(h)阶,其中h是数值方法的时间步长。我们提出的3种估计方法均比文献中已有的EM-MLE方法更精确。
[1] | Papaspiliopoulos O, Beskos A. An introduction to modelling and likelihood inference with stochastic differential equations. Great Britain:Warwick, 2011. |
[2] | Morel J M, Takens F, Teissier B. Parameter estimation in stochastic differential equations[M]. Berlin Heidelberg:Springer-Verlag, 2008:6-10. |
[3] | Breton A L. On continuous and discrete sampling for parameter estimation in diffusion type processes[M]. Berlin Heidelberg:Springer, 1976:124-144. |
[4] | Dorogovcev A J. The asymptotic properties of least square estimators for regression coefficients[J]. Problemy Peredai Informacii, 1973, 4:49-57. |
[5] | Pedersen A R. A new approach to maximum likelihood estimation for stochatic differential equations based on discrete observations[J]. Scandinavian Journal of Statistics, 1995, 22(1):55-71. |
[6] | Milstein G N, Tretyakov M V. Stochastic numerics for mathematical physics[M]. Berlin Heidelberg:Springer, 2004:9-10. |
[7] | Prakasa R. On Bayes estimation for diffusion fields[M]. North Holland:Statistical Publishing Society, 1984:575-590. |
[8] | Kloeden P E, Platen E. Numerical solution of stochastic differential equations[M]. Berlin Heidelberg:Springer-Verlag, 1992:305. |