Meta Optimality for Demographic Parity Constrained Regression via Post-Processing
By: Kazuto Fukuchi
Potential Business Impact:
Makes computer predictions fair for everyone.
We address the regression problem under the constraint of demographic parity, a commonly used fairness definition. Recent studies have revealed fair minimax optimal regression algorithms, the most accurate algorithms that adhere to the fairness constraint. However, these analyses are tightly coupled with specific data generation models. In this paper, we provide meta-theorems that can be applied to various situations to validate the fair minimax optimality of the corresponding regression algorithms. Furthermore, we demonstrate that fair minimax optimal regression can be achieved through post-processing methods, allowing researchers and practitioners to focus on improving conventional regression techniques, which can then be efficiently adapted for fair regression.
Similar Papers
Explainable post-training bias mitigation with distribution-based fairness metrics
Machine Learning (CS)
Makes AI fair without retraining.
Set to Be Fair: Demographic Parity Constraints for Set-Valued Classification
Machine Learning (Stat)
Makes AI fairer by avoiding biased predictions.
General Post-Processing Framework for Fairness Adjustment of Machine Learning Models
Machine Learning (CS)
Makes computer decisions fair without retraining.