Fairness for Robust Log Loss Classification
This work addresses fairness in classification for social applications, presenting an incremental improvement by incorporating fairness into a robust optimization framework.
The authors tackled the problem of developing classification methods that achieve high accuracy while ensuring fairness across different groups by deriving a new classifier from distributional robustness principles, which resulted in a parametric exponential family conditional distribution resembling truncated logistic regression and demonstrated practical advantages on three benchmark fairness datasets.
Developing classification methods with high accuracy that also avoid unfair treatment of different groups has become increasingly important for data-driven decision making in social applications. Many existing methods enforce fairness constraints on a selected classifier (e.g., logistic regression) by directly forming constrained optimizations. We instead re-derive a new classifier from the first principles of distributional robustness that incorporates fairness criteria into a worst-case logarithmic loss minimization. This construction takes the form of a minimax game and produces a parametric exponential family conditional distribution that resembles truncated logistic regression. We present the theoretical benefits of our approach in terms of its convexity and asymptotic convergence. We then demonstrate the practical advantages of our approach on three benchmark fairness datasets.