Auditing and Enforcing Conditional Fairness via Optimal Transport
This addresses fairness in machine learning for applications requiring conditional fairness, though it is incremental as it builds on existing demographic parity techniques.
The paper tackles the problem of auditing and enforcing conditional demographic parity (CDP) in predictive models, particularly when conditioning variables have many levels or outputs are continuous, by proposing novel measures based on optimal transport and regularization methods, achieving full equality of conditional distributions in validation on real-world datasets.
Conditional demographic parity (CDP) is a measure of the demographic parity of a predictive model or decision process when conditioning on an additional feature or set of features. Many algorithmic fairness techniques exist to target demographic parity, but CDP is much harder to achieve, particularly when the conditioning variable has many levels and/or when the model outputs are continuous. The problem of auditing and enforcing CDP is understudied in the literature. In light of this, we propose novel measures of {conditional demographic disparity (CDD)} which rely on statistical distances borrowed from the optimal transport literature. We further design and evaluate regularization-based approaches based on these CDD measures. Our methods, \fairbit{} and \fairlp{}, allow us to target CDP even when the conditioning variable has many levels. When model outputs are continuous, our methods target full equality of the conditional distributions, unlike other methods that only consider first moments or related proxy quantities. We validate the efficacy of our approaches on real-world datasets.