Amortized Equation Discovery in Hybrid Dynamical Systems

Official code of Amortized Equation Discovery in Hybrid Dynamical Systems in ICML 2024

🚀 Performance

AMORE demonstrates remarkable proficiency in handling complex segmentation tasks that require: (a) reasoning based on world knowledge; (b) inference of future events; and (c) a comprehensive understanding of video content.

🛠️ Installation

pip install -r requirements.txt

🦄 Training and Validation

1. Training Data Preparation

Before training, please download the datasets and put them in data. The Mass-spring Hopping dataset has already been put in the folder, which is a good start.

2. Training and evaluating AMORE

python run-amore.py --config "./configs/spring-mass.yaml" --device "cuda:0"

⭐ Cite

If you find this project useful in your research, please consider citing:

@article{liu2024amortized,
  title={Amortized Equation Discovery in Hybrid Dynamical Systems},
  author={Liu, Yongtuo and Magliacane, Sara and Kofinas, Miltiadis and Gavves, Efstratios},
  journal={arXiv preprint arXiv:2406.03818},
  year={2024}
}

🎖️ Acknowledgement

This work is built upon the SNLDS, REDSDS, and PySINDy.