A Unified Post-Processing Framework for Group Fairness in Classification
This work addresses fairness in machine learning for classification tasks, offering a unified solution that is incremental in building upon existing post-processing methods.
The authors tackled the problem of ensuring group fairness in classification by introducing a post-processing algorithm that unifies multiple fairness criteria and works for multiclass problems. Their method, LinearPost, achieves fairness by linearly transforming predictions and shows empirical advantages in high fairness regimes compared to existing approaches.
We present a post-processing algorithm for fair classification that covers group fairness criteria including statistical parity, equal opportunity, and equalized odds under a single framework, and is applicable to multiclass problems in both attribute-aware and attribute-blind settings. Our algorithm, called "LinearPost", achieves fairness post-hoc by linearly transforming the predictions of the (unfair) base predictor with a "fairness risk" according to a weighted combination of the (predicted) group memberships. It yields the Bayes optimal fair classifier if the base predictors being post-processed are Bayes optimal, otherwise, the resulting classifier may not be optimal, but fairness is guaranteed as long as the group membership predictor is multicalibrated. The parameters of the post-processing can be efficiently computed and estimated from solving an empirical linear program. Empirical evaluations demonstrate the advantage of our algorithm in the high fairness regime compared to existing post-processing and in-processing fair classification algorithms.