An Intersectional Definition of Fairness

By James R. Foulds, Shimei Pan. 2018

231 citations recorded by the source corpus.

We propose differential fairness, a multi-attribute definition of fairness in machine learning which is informed by intersectionality, a critical lens arising from the humanities literature, leveraging connections between differential privacy and legal notions of fairness. We show that our criterion behaves sensibly for any subset of the set of protected attributes, and we prove economic, privacy, and generalization guarantees. We provide a learning algorithm which respects our differential fairness criterion. Experiments on the COMPAS criminal recidivism dataset and census data demonstrate the utility of our methods.

Subfields: Artificial Intelligence, Safety Research, Sociology and Political Science

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Safety Research

In Sociology and Political Science

Citations in the corpus

50 Years of Test (Un)fairness: Lessons for Machine Learning (2018, background): “ality can be modeled. This might include, for example, how to quantify fairness along one dimension (e.g., age) conditioned on another dimension (e.g., skin tone), as recent work has begun to address [27, 39]. 6 CONCLUSIONS The spike in interest in test fairness in the 1960s arose during a time of s”
Directional Bias Amplification (2021, background): “It is the result of the algorithm (Foulds et al., 2018), and unlike some other forms of bias, cannot be solely attributed to the dataset.”
A Systematic Study of Bias Amplification (2022, background): “In line with prior work [14, 43, 47], we find that models trained for this classification task, indeed, amplify biases present in their training data.”
Towards Intersectionality in Machine Learning: Including More Identities, Handling Underrepresentation, and Performing Evaluation (2022, methodology): “To generalize the equal opportunity measure to more than two groups, we adopt a method similar to prior work [40, 77, 121].”
Awareness in practice: tensions in access to sensitive attribute data for antidiscrimination (2019, background): “[17, 32, 39, 40, 48, 50, 65]. When labeled data are not available, researchers have made powerful discoveries by augmenting existing data through the inference or construction of labeled data de novo [5, 10, 27, 46]. At times, this work has insufficiently acknowledged the full range of challenges to”
Bayesian Modeling of Intersectional Fairness: The Variance of Bias (2018, methodology): “A simple baseline, proposed by [13] to address the zero count issue, is to put a Dirichlet prior on the probabilities in Equation 2.”