A model’s tendency to give unbalanced or unfair answers, often inherited from the data it was trained on.
← Back to the full glossary