Do I need to specify any labels in my data to detect anomalies?
Updated
No, the Anomaly Detector does not require any labels in your data. The anomaly scores are automatically generated by the system. BigML generally considers anomaly scores greater than 0.60 as anomalous instances.