α-stable运动驱动的OU过程的拟似然估计  

Quasi-Likelihood Estimation for OU Processes Driven by α-Stable Motions

作　　者：方龙祥[1] 张新生[2] 

机构地区：[1]安徽师范大学数学计算机科学学院,芜湖241003 [2]复旦大学管理学院统计系,上海200433

出　　处：《数学学报（中文版）》2014年第2期395-408,共14页Acta Mathematica Sinica：Chinese Series

基　　金：国家自然科学基金资助项目(11071045;11201003);安徽省自然科学基金(1408085MA07;1208085MA11);高校省级重点科学研究项目(KJ2013A137)

摘　　要：对于α-stable运动驱动的OU过程,其转移密度函数没有明确的显式表达式,但是条件特征函数已知.本文基于离散观测样本,借助条件特征函数,研究了α-stable运动驱动的OU过程的拟似然估计,并证明了拟似然估计量的相合性和渐近有效性.模拟显示所提出的估计方法是准确和稳定的.For OU processes driven by α-stable motions, its transition density func tion has no explicit form, while the conditional characteristic function is known. In this paper, we present an quasi-likelihood estimation procedure for parameters of the discretely sampled process of OU processes driven by s-stable motions. The proposedprocedure is based on the conditional characteristic function, and the quasi-likelihood estimator is proved to be consistent and asymptotically efficient under some mild con ditions. Finally, Monte Carlo simulations are conducted to demonstrate the accuracy and stability of proposed estimator.

关 键 词：α-stable运动驱动的OU过程 条件特征函数 拟似然估计 FISHER信息阵 

分 类 号：O211.4[理学—概率论与数理统计] O211.62[理学—数学]