This project was initially described in the full research track paper Spatio-temporal Diffusion Point Processes at KDD 2023 in Long Beach, CA. Contributors to this project are from the Future Intelligence laB (FIB) at Tsinghua University.
The code is tested under a Linux desktop with torch 1.7 and Python 3.7.10.
- Tested OS: Linux
- Python >= 3.7
- PyTorch == 1.7.1
- Tensorboard
- Install PyTorch 1.7.1 with the correct CUDA version.
- Use the
pip install -r requirements. txt
command to install all of the Python modules and packages used in this project.
Use the following command to train DSTPP on Earthquake
dataset:
python app.py --dataset Earthquake --mode train --timesteps 500 --samplingsteps 500 --batch_size 64 --total_epochs 2000
To train DSTPP on other datasets:
python app.py --dataset COVID19 --mode train --timesteps 500 --samplingsteps 500 --batch_size 64 --total_epochs 2000
python app.py --dataset Citibike --mode train --timesteps 500 --samplingsteps 500 --batch_size 128 --total_epochs 2000
python app.py --dataset Independent --mode train --timesteps 500 --samplingsteps 500 --batch_size 128 --total_epochs 2000
The trained models are saved in ModelSave/
.
The logs are saved in logs/
.
The implemention is based on DDPM.
If you found this library useful in your research, please consider citing:
@inproceedings{yuan2023DSTPP,
author = {Yuan, Yuan and Ding, Jingtao and Shao, Chenyang and Jin, Depeng and Li, Yong},
title = {Spatio-Temporal Diffusion Point Processes},
year = {2023},
booktitle = {Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining},
pages = {3173–3184},
}