Supervision

class dtaianomaly.anomaly_detection.Supervision(*values)

Supervision types.

An enum for the different supervision types for anomaly detectors. Valid supervision types are:

• Unsupervised: the anomaly detector does not need any labels or training data.

• Semi-supervised: The anomaly detector requires normal training data, but no training labels.

• Supervised: The anomaly detector requires both training data and training labels. The training data may contain anomalies.