Statistical Guarantees for Fairness Aware Plug-In Algorithms
This work addresses statistical reliability for fairness-aware classification, with incremental improvements to existing methods.
The paper tackles the lack of statistical guarantees for a fairness-aware plug-in algorithm in binary classification, proving its consistency and deriving finite sample guarantees, and also proposes a protocol to ensure both fairness and differential privacy for sensitive features.
A plug-in algorithm to estimate Bayes Optimal Classifiers for fairness-aware binary classification has been proposed in (Menon & Williamson, 2018). However, the statistical efficacy of their approach has not been established. We prove that the plug-in algorithm is statistically consistent. We also derive finite sample guarantees associated with learning the Bayes Optimal Classifiers via the plug-in algorithm. Finally, we propose a protocol that modifies the plug-in approach, so as to simultaneously guarantee fairness and differential privacy with respect to a binary feature deemed sensitive.