Fairness-aware Regression Robust to Adversarial Attacks
This addresses the need for fairness-aware machine learning that can withstand adversarial manipulations, though it is incremental as it builds on existing minimax frameworks.
The paper tackles the problem of designing fair regression models robust to adversarial attacks, achieving better performance on poisoned datasets compared to other fair models in both prediction accuracy and group-based fairness measures.
In this paper, we take a first step towards answering the question of how to design fair machine learning algorithms that are robust to adversarial attacks. Using a minimax framework, we aim to design an adversarially robust fair regression model that achieves optimal performance in the presence of an attacker who is able to add a carefully designed adversarial data point to the dataset or perform a rank-one attack on the dataset. By solving the proposed nonsmooth nonconvex-nonconcave minimax problem, the optimal adversary as well as the robust fairness-aware regression model are obtained. For both synthetic data and real-world datasets, numerical results illustrate that the proposed adversarially robust fair models have better performance on poisoned datasets than other fair machine learning models in both prediction accuracy and group-based fairness measure.