Code for the experiments in our preprint:

"Sample Complexity of Bias Detection with Subsampled Point-to-Subspace Distances"

about Fairness in AI, in which we propose a method for which we bound the number of samples required (=Sample Complexity) to spot the bias present in some data for a certain group to a given probability. We test our theoretical results using the Adult dataset and data from the folktables.

Reference

If you like the work, please cite it:

@misc{matilla2025samplecomplexitybiasdetection,
      title={Sample Complexity of Bias Detection with Subsampled Point-to-Subspace Distances},
      author={M. Matilla, Germán and Mareček, Jakub},
      year={2025},
      eprint={2502.02623},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2502.02623v1},
}