Explainable AI (XAI) in Anomaly Detection.

Explaining why a point was flagged[cite: 1].

In fraud or medical cases, 'The black box said so' is not enough. Use SHAP or LIME to identify the features that contributed most to the anomaly score[cite: 1].