arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2404.05554math.STmath.PRstat.TH
收藏

遍历 Volterra Ornstein-Uhlenbeck 过程中的最大似然估计

Maximum likelihood estimation in the ergodic Volterra Ornstein-Uhlenbeck process

Mohamed Ben Alaya, Martin Friesen, Jonas Kremer

AI总结:

本文研究遍历 Volterra Ornstein-Uhlenbeck 过程的漂移参数推断,分别在连续、高频和低频观测下建立 MLE 与矩估计的渐近性质。

摘要:

我们研究了遍历情形下定义于 R 上的 Volterra Ornstein-Uhlenbeck 过程漂移参数的统计推断。对于连续时间观测,我们推导了相应的最大似然估计量,并证明其在参数上局部一致地具有强相合性与渐近正态性。对于离散高频观测情形,我们通过对连续时间最大似然估计量进行离散化,证明了类似结果。最后,对于离散低频观测,我们证明矩估计法具有相合性。我们的证明关键依赖于大数定律。

英文摘要:

We study statistical inference of the drift parameters for the Volterra Ornstein-Uhlenbeck process on R in the ergodic regime. For continuous-time observations, we derive the corresponding maximum likelihood estimators and show that they are strongly consistent and asymptotically normal locally uniformly in the parameters. For the case of discrete high-frequency observations, we prove similar results by discretization of the continuous-time maximum likelihood estimator. Finally, for discrete low-frequency observations, we show that the method of moments is consistent. Our proofs are crucially based on the law of large numbers.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑