Provable Data Subset Selection for Efficient Neural Network Training [Accepted to ICML'23]

1 DataHeroes | 2 Rice University | 3 CSAIL, MIT | 4 University of Haifa

Murad Tukan, Samson Zhou, Alaa Maalouf, Daniela Rus, Vladimir Braverman, and Dan Feldman

We kindly refer you to our paper:

• This is the first coreset paper for Radial Basis Function Neural Networks (RBFNN).
• This coreset can serve a large family of functions that can be represented by any RBFNN model.
• Finally, our coreset can be used for the task of subset selection in the context of training deep neural networks.

In this repository, you will find our coreset construction code (tested with Python 3.9).

Citation

If you find this work helpful please cite us:

@article{tukan2023provable,

  title={Provable Data Subset Selection For Efficient Neural Network Training},
  
  author={Tukan, Murad and Zhou, Samson and Maalouf, Alaa and Rus, Daniela and Braverman, Vladimir and Feldman, Dan},
  
  journal={arXiv preprint arXiv:2303.05151},
  
  year={2023}
}