In this paper, we are interested in linear prediction of a particular kind of stochastic process, namely a marked temporal point process. The observations are event times recorded on the real line, with marks attached to each event. We show that in this case, linear prediction generalises straightforwardly from the theory of prediction for stationary stochastic processes. We propose two recursive methods to solve the linear prediction problem and show that these are computationally efficient. The first relies on a Wiener-Hopf integral equation and a corresponding set of differential equations. It is particularly well-adapted to autoregressive processes. In the second method, we develop an innovations algorithm tailored for moving-average processes. Both methods are assessed by an extensive simulation study and applied to a real-world dataset of polling data ahead of the 2022 French elections. In a particular case, we extend the "model independent origin" idea of Jaisson (2015) to the marked Hawkes process through its autoregressive representation. As a corollary, we also improve on existing non-recursive estimators such as that proposed by Bacry and Muzy (2016).
翻译:在本文中,我们有兴趣对某种特定的随机过程进行线性预测,即一个标志性的时点过程。观测是记录在真实线上的事件时间,每个事件都有标记。我们显示,在这种情况下,线性预测直接地从静止随机过程的预测理论中归纳出来。我们建议了两种循环方法来解决线性预测问题,并表明这些是计算效率高的。第一种方法依赖于维纳-Hopf综合方程式和一套相应的差别方程式。它特别适合自动递进过程。第二种方法是,我们开发一种适应移动平均过程的创新算法。两种方法都通过广泛的模拟研究加以评估,并应用于2022年法国选举之前的民意数据真实世界数据集。在一种特定案例中,我们将Jaisson(2015年)的“模型独立来源”概念通过自动递增的表达方式扩大到有标记的霍克斯进程。作为推论的推论,我们还改进了巴克利和穆兹(2016年)提议的现有的非封闭式估测算器。