AutoSTPP
Automatic Integration for Neural Spatio-Temporal Point Process models (AI-STPP) is a new paradigm for exact, efficient, non-parametric inference of point process. It is capable of learning complicated underlying intensity functions, like a damped sine wave.
Install / Use
/learn @Rose-STL-Lab/AutoSTPPREADME
<p align="center" >
<a href="https://github.com/Rose-STL-Lab/AI-STPP"><img src="https://raw.githubusercontent.com/Rose-STL-Lab/AutoSTPP/refs/heads/main/Auto-STPP.png" width="256" height="256" alt="AI-STPP"></a>
</p>
<h1 align="center">Auto-STPP</h1>
<h4 align="center">✨Automatic Integration for Neural Spatiotemporal Point Process✨</h4>
<p align="center">
<a href="https://raw.githubusercontent.com/Rose-STL-Lab/AutoSTPP/refs/heads/main/LICENSE"><img src="https://img.shields.io/badge/License-MIT-yellow.svg" alt="license"></a>
<img src="https://img.shields.io/badge/Python-3.10+-yellow" alt="python">
<img src="https://img.shields.io/badge/Version-1.1.0-green" alt="version">
</p>
| Introduction
Automatic Integration for Neural Spatio-Temporal Point Process models (Auto-STPP) is a new paradigm for exact, efficient, non-parametric inference of spatiotemporal point process.
| Citation
[2310.06179] Automatic Integration for Spatiotemporal Neural Point Processes
@article{zhou2023automatic,
title={Automatic Integration for Spatiotemporal Neural Point Processes},
author={Zhou, Zihao and Yu, Rose},
journal={arXiv preprint arXiv:2310.06179},
year={2023}
}
| Installation
Dependencies: make, conda-lock
make create_environment
conda activate autoint-stpp
| Dataset Download
python src/download_data.py
| Training and Testing
Specify the parameters in configs/autoint_stpp.yaml and then run
make run_stpp config=autoint_stpp
