POTs: protective optimization technologies

@article{Kulynych2020POTsPO,
  title={POTs: protective optimization technologies},
  author={Bogdan Kulynych and R. Overdorf and C. Troncoso and S. G{\"u}rses},
  journal={Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency},
  year={2020}
}
  • Bogdan Kulynych, R. Overdorf, +1 author S. Gürses
  • Published 2020
  • Computer Science
  • Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency
  • Algorithmic fairness aims to address the economic, moral, social, and political impact that digital systems have on populations through solutions that can be applied by service providers. Fairness frameworks do so, in part, by mapping these problems to a narrow definition and assuming the service providers can be trusted to deploy countermeasures. Not surprisingly, these decisions limit fairness frameworks' ability to capture a variety of harms caused by systems. We characterize fairness… CONTINUE READING
    18 Citations

    Figures, Tables, and Topics from this paper

    Questioning the assumptions behind fairness solutions
    • 13
    • PDF
    Politics of Adversarial Machine Learning
    • 3
    • PDF
    Poisoning Attacks on Algorithmic Fairness
    • 7
    • PDF
    Attacking Machine Learning Models for Social Good
    • PDF
    Evading classifiers in discrete domains with provable optimality guarantees
    • 9
    • PDF
    Aligning AI Optimization to Community Well-Being
    • Jonathan Stray
    • Computer Science
    • International Journal of Community Well-Being
    • 2020
    • PDF

    References

    SHOWING 1-3 OF 3 REFERENCES
    Concrete Problems in AI Safety
    • 802
    • Highly Influential
    • PDF
    Shortest Path Network Interdiction With Goal Threshold
    • 8
    • Highly Influential
    Shortest-path network interdiction
    • 429
    • Highly Influential
    • PDF