Pairwise Fairness for Ordinal Regression
Abstract
We initiate the study of fairness for ordinal regression. We adapt two fairness notions previously considered in fair ranking and propose a strategy for training a predictor that is approximately fair according to either notion. Our predictor has the form of a threshold model, composed of a scoring function and a set of thresholds, and our strategy is based on a reduction to fair binary classification for learning the scoring function and local search for choosing the thresholds. We provide generalization guarantees on the error and fairness violation of our predictor, and we illustrate the effectiveness of our approach in extensive experiments.
Cite
Text
Kleindessner et al. "Pairwise Fairness for Ordinal Regression." Artificial Intelligence and Statistics, 2022.Markdown
[Kleindessner et al. "Pairwise Fairness for Ordinal Regression." Artificial Intelligence and Statistics, 2022.](https://mlanthology.org/aistats/2022/kleindessner2022aistats-pairwise/)BibTeX
@inproceedings{kleindessner2022aistats-pairwise,
title = {{Pairwise Fairness for Ordinal Regression}},
author = {Kleindessner, Matthäus and Samadi, Samira and Bilal Zafar, Muhammad and Kenthapadi, Krishnaram and Russell, Chris},
booktitle = {Artificial Intelligence and Statistics},
year = {2022},
pages = {3381-3417},
volume = {151},
url = {https://mlanthology.org/aistats/2022/kleindessner2022aistats-pairwise/}
}