An efficient method for federated classification in hyperbolic spaces, as introduced in our paper:
For the single-cell embeddings used in our paper, please refer to sc-mvae.
Please see the shell files for more details.
Please contact Saurav Prakash (sauravp2@illinois.edu) and Chao Pan (chaopan2@illinois.edu) if you have any questions.
Please consider citing our paper if you find our repository useful!
@article{
prakash2024_hyperbolic_fl,
title={Federated Classification in Hyperbolic Spaces via Secure Aggregation of Convex Hulls},
author={Prakash, Saurav and Sima, Jin and Pan, Chao and Chien, Eli and Milenkovic, Olgica},
journal={Transactions on Machine Learning Research},
year={2023},
url={https://openreview.net/forum?id=umggDfMHha}
}
This repository also uses the Poincare-SVM algorithm from our previous paper
@article{pan2023provably,
title={Provably accurate and scalable linear classifiers in hyperbolic spaces},
author={Pan, Chao and Chien, Eli and Tabaghi, Puoya and Peng, Jianhao and Milenkovic, Olgica},
journal={Knowledge and Information Systems},
pages={1--34},
year={2023},
publisher={Springer}
}
@inproceedings{chien2021highly,
title={Highly Scalable and Provably Accurate Classification in Poincar{\'e} Balls},
author={Chien, Eli and Pan, Chao and Tabaghi, Puoya and Milenkovic, Olgica},
booktitle={2021 IEEE International Conference on Data Mining (ICDM)},
pages={61--70},
year={2021},
organization={IEEE}
}