Signatures to help interpretability of anomalies
This addresses interpretability issues for astronomers in anomaly detection, but appears incremental as it builds on existing concepts for feature attribution.
The paper tackles the problem of interpretability in anomaly detection by introducing anomaly signatures to highlight features contributing to decisions, aiming to help astronomers understand why events are tagged as anomalies.
Machine learning is often viewed as a black box when it comes to understanding its output, be it a decision or a score. Automatic anomaly detection is no exception to this rule, and quite often the astronomer is left to independently analyze the data in order to understand why a given event is tagged as an anomaly. We introduce here idea of anomaly signature, whose aim is to help the interpretability of anomalies by highlighting which features contributed to the decision.