Path-Specific Counterfactual Fairness

AAAI 2019 pp. 7801-7808

doi:10.1609/AAAI.V33I01.33017801 /aaai/2019/chiappa2019aaai-path/

Abstract

We consider the problem of learning fair decision systems from data in which a sensitive attribute might affect the decision along both fair and unfair pathways. We introduce a counterfactual approach to disregard effects along unfair pathways that does not incur in the same loss of individual-specific information as previous approaches. Our method corrects observations adversely affected by the sensitive attribute, and uses these to form a decision. We leverage recent developments in deep learning and approximate inference to develop a VAE-type method that is widely applicable to complex nonlinear models.

PDF AAAI Semantic Scholar

Cite

Text

Chiappa. "Path-Specific Counterfactual Fairness." AAAI Conference on Artificial Intelligence, 2019. doi:10.1609/AAAI.V33I01.33017801

Markdown

[Chiappa. "Path-Specific Counterfactual Fairness." AAAI Conference on Artificial Intelligence, 2019.](https://mlanthology.org/aaai/2019/chiappa2019aaai-path/) doi:10.1609/AAAI.V33I01.33017801

BibTeX

@inproceedings{chiappa2019aaai-path,
  title     = {{Path-Specific Counterfactual Fairness}},
  author    = {Chiappa, Silvia},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2019},
  pages     = {7801-7808},
  doi       = {10.1609/AAAI.V33I01.33017801},
  url       = {https://mlanthology.org/aaai/2019/chiappa2019aaai-path/}
}