no code implementations • 2 May 2023 • Aki Barry, Lei Han, Gianluca Demartini
By adding noise to the original datasets, we can explore the relationship between the quality of the training data and the fairness of the output of the models trained on that data.