When running an unsupervised anomaly detection model, the machine learning algorithm requires no labeled data. When training the model, the target_column_name parameter must be set to NULL.

Semi-supervised learning for anomaly detection uses a specific set of labeled data along with unlabeled data to detect anomalies. To enable this, use the experimental and semisupervised options. The target_column_name parameter must specify a column whose only allowed values are 0 (normal), 1 (anomalous), and NULL (unlabeled). All rows are used to train the unsupervised component, while the rows with a value different than NULL are used to train the supervised component.