Bellman Meets Hawkes: Model-Based Reinforcement Learning via Temporal Point Processes

The repo for Bellman Meets Hawkes: Model-Based Reinforcement Learning via Temporal Point Processes

This is the demonstration of 8 events synthetic simulation in the paper.

Quick Start

1. To run the model, simply run the main file:

python main.py

2. The config file can be found in:

configs/config.yaml

3. The simulation is implemented by:

env_hawkes_event.py

Reference

If you use or extend our work, please cite the following paper:

@article{qu2022bellman,
  title={Bellman Meets Hawkes: Model-Based Reinforcement Learning via Temporal Point Processes},
  author={Qu, Chao and Tan, Xiaoyu and Xue, Siqiao and Shi, Xiaoming and Zhang, James and Mei, Hongyuan},
  journal={arXiv preprint arXiv:2201.12569},
  year={2022}
}