The widespread use of automated decision processes in many areas of our society raises serious ethical issues with respect to the fairness of the process and the possible resulting discrimination. To solve this issue, we propose a novel adversarial training approach called GANSan for learning a sanitizer whose objective is to prevent the possibility of any discrimination (i.e., direct and indirect) based on a sensitive attribute by removing the attribute itself as well as the existing correlations with the remaining attributes. Our method GANSan is partially inspired by the powerful framework of generative adversarial networks (in particular Cycle-GANs), which offers a flexible way to learn a distribution empirically or to translate between two different distributions. In contrast to prior work, one of the strengths of our approach is that the sanitization is performed in the same space as the original data by only modifying the other attributes as little as possible, thus preserving the interpretability of the sanitized data. Consequently, once the sanitizer is trained, it can be applied to new data locally by an individual on their profile before releasing it. Finally, experiments on real datasets demonstrate the effectiveness of the approach as well as the achievable trade-off between fairness and utility.

This paper presents an approach to overapproximate the reachable set of states of a system whose uncertainties are arbitrarily time-varying. Most approaches generally assume piecewise continuity or sometimes Riemannintegrability of the uncertainties. In this paper we go one step further, only assuming Lebesgue measurability, which is the weakest meaningful hypothesis. We develop our new technique, based on a decomposition of components as a difference of positive functions, for separable systems, a generalization of control-affine systems. We compare the overapproximation produced by our method with the ones obtained using the tools Flow* and CORA on simple examples, and show that correct outer-approximations of the reachable sets are computable with a high degree of precision even for these general forms of uncertainties.