Flexible marked spatio-temporal point processes with applications to event sequences from association football
提出一种新的标记点过程模型，将标记霍克斯过程的特性聚焦于标记空间，允许单独建模事件发生时间，并利用贝叶斯框架整合协变量信息，应用于足球比赛事件序列，推断比赛动态、球队能力并预测进球或犯规等事件。
Abstract We develop a new family of marked point processes by focusing the characteristic properties of marked Hawkes processes exclusively on the space of marks, allowing a separate model specification for the occurrence times. We develop a Bayesian framework for their inference and prediction that can naturally accommodate covariate information to drive cross-excitations, offering broad flexibility for real-world applications. The framework is applied to in-game event sequences from association football, resulting in inferences about previously unquantified characteristics of game dynamics, extraction of event-specific team abilities and predictions for event occurrences, such as goals or fouls in a specified interval of time.