Debiased Imitation Learning for Modulated Temporal Point Processes
Document Type
Conference Proceeding
Publication Date
1-1-2023
Abstract
Temporal event sequences associated with different event types (e.g., location indices, disease types) are observed in various applications such as disaster resilience, criminology, and healthcare. Temporal point processes (TPPs) have been developed to capture the exciting patterns between events and forecast future events quantitatively. Unfortunately, the events with different types often suffer from unknown biased observations in real-world scenarios due to external interference. Accordingly, the temporal point processes learned by conventional maximum likelihood estimation (MLE) from such biased data may be misspecified and may lead to inaccurate predictions. To overcome this issue, we model biased event sequences as modulating TPPs with additional unknown thinning processes. Furthermore, we develop a novel debiased imitation learning framework to learn the modulated TPPs and suppress the negative influences of biased data, which is more robust than conventional MLE. When applying the debiased imitation learning framework, we design a simple but effective reward function based on the historical embedding obtained by the TPP model. Experiments on three real-world datasets demonstrate that our proposed method significantly outperforms existing methods.
Publication Source (Journal or Book title)
2023 SIAM International Conference on Data Mining, SDM 2023
First Page
460
Last Page
468
Recommended Citation
Li, Z., Zhou, Z., Sun, M., & Xu, H. (2023). Debiased Imitation Learning for Modulated Temporal Point Processes. 2023 SIAM International Conference on Data Mining, SDM 2023, 460-468. Retrieved from https://repository.lsu.edu/eecs_pubs/2703
