Variable Selection in the Context of AI Fairness
The findings suggest that the exclusion of sensitive or critical variables may compromise equity between subgroups and retaining all relevant variables could reduce implicit bias, so the interdisciplinary approach could provide deeper insight into the ethical implications and compliance with regulatory standards.