Fair Supervised Learning with A Simple Random Sampler of Sensitive Attributes
This work addresses fairness in machine learning for industrial applications, offering a more flexible and efficient approach than existing methods, though it appears incremental in nature.
The paper tackled the problem of fairness-aware supervised learning by proposing a neural network-based fairness penalty with a simple random sampler of sensitive attributes, enabling handling of versatile attribute formats and achieving better utility and fairness measures on benchmark datasets than competing methods.
As the data-driven decision process becomes dominating for industrial applications, fairness-aware machine learning arouses great attention in various areas. This work proposes fairness penalties learned by neural networks with a simple random sampler of sensitive attributes for non-discriminatory supervised learning. In contrast to many existing works that critically rely on the discreteness of sensitive attributes and response variables, the proposed penalty is able to handle versatile formats of the sensitive attributes, so it is more extensively applicable in practice than many existing algorithms. This penalty enables us to build a computationally efficient group-level in-processing fairness-aware training framework. Empirical evidence shows that our framework enjoys better utility and fairness measures on popular benchmark data sets than competing methods. We also theoretically characterize estimation errors and loss of utility of the proposed neural-penalized risk minimization problem.