Preserving Fairness in AI under Domain Shift

Open in new window