Neural Spectral Marked Point Processes
Neural Spectral Marked Point Processes
复制标题
DOI：
--
复制
发表时间：
2021-06
期刊：
影响因子：
--
通讯作者：
Shixiang Zhu;Haoyun Wang;Xiuyuan Cheng;Yao Xie
中科院分区：
文献类型：
--
作者：
Shixiang Zhu;Haoyun Wang;Xiuyuan Cheng;Yao Xie
自激和互激点过程是机器学习和相关离散事件数据统计中的流行模型。迄今为止，大多数现有的模型假设平稳内核（包括经典的霍克斯过程）和简单的参数模型。具有复杂事件数据的现代应用需要更通用的点过程模型，除了时间和位置信息之外，该模型还可以包含事件的上下文信息（称为标记）。此外，这样的应用通常需要非平稳模型来捕获更复杂的时空依赖性。为了应对这些挑战，一个关键的问题是设计一个通用的影响核的点过程模型。在本文中，我们介绍了一种新的和一般的基于神经网络的非平稳影响内核具有高表现力的处理复杂的离散事件数据，同时提供理论性能保证。我们证明了上级性能相比，我们所提出的方法的国家的最先进的合成和真实的数据。
Self- and mutually-exciting point processes are popular models in machine learning and statistics for dependent discrete event data. To date, most existing models assume stationary kernels (including the classical Hawkes processes) and simple parametric models. Modern applications with complex event data require more general point process models that can incorporate contextual information of the events, called marks, besides the temporal and location information. Moreover, such applications often require non-stationary models to capture more complex spatio-temporal dependence. To tackle these challenges, a key question is to devise a versatile influence kernel in the point process model. In this paper, we introduce a novel and general neural network-based non-stationary influence kernel with high expressiveness for handling complex discrete events data while providing theoretical performance guarantees. We demonstrate the superior performance of our proposed method compared with the state-of-the-art on synthetic and real data.