Fair Meta-Learning: Learning How to Learn Fairly
It addresses fairness in meta-learning for practitioners dealing with data scarcity or bias, but it is incremental as it builds on existing MAML methods.
The paper tackles the problem of training fair machine learning models with limited or biased data by adapting the MAML algorithm to include a fairness regularization term, resulting in Fair-MAML, which enables training from few examples with related task data and shows empirical improvements over baselines.
Data sets for fairness relevant tasks can lack examples or be biased according to a specific label in a sensitive attribute. We demonstrate the usefulness of weight based meta-learning approaches in such situations. For models that can be trained through gradient descent, we demonstrate that there are some parameter configurations that allow models to be optimized from a few number of gradient steps and with minimal data which are both fair and accurate. To learn such weight sets, we adapt the popular MAML algorithm to Fair-MAML by the inclusion of a fairness regularization term. In practice, Fair-MAML allows practitioners to train fair machine learning models from only a few examples when data from related tasks is available. We empirically exhibit the value of this technique by comparing to relevant baselines.