An Algorithmic Framework for Fairness Elicitation
This addresses fairness in machine learning for stakeholders needing nuanced fairness beyond simple statistical definitions, though it is incremental as it builds on existing fairness elicitation methods.
The paper tackles the problem of learning fair models when fairness definitions are complex and require stakeholder input, by introducing a framework for eliciting pairwise fairness constraints and providing a provably convergent algorithm that learns accurate models subject to these constraints, with preliminary results on the COMPAS dataset.
We consider settings in which the right notion of fairness is not captured by simple mathematical definitions (such as equality of error rates across groups), but might be more complex and nuanced and thus require elicitation from individual or collective stakeholders. We introduce a framework in which pairs of individuals can be identified as requiring (approximately) equal treatment under a learned model, or requiring ordered treatment such as "applicant Alice should be at least as likely to receive a loan as applicant Bob". We provide a provably convergent and oracle efficient algorithm for learning the most accurate model subject to the elicited fairness constraints, and prove generalization bounds for both accuracy and fairness. This algorithm can also combine the elicited constraints with traditional statistical fairness notions, thus "correcting" or modifying the latter by the former. We report preliminary findings of a behavioral study of our framework using human-subject fairness constraints elicited on the COMPAS criminal recidivism dataset.