Which algorithm does AI Studio use for Anomaly Detection?

AI Studio uses the Isolation Forest algorithm to detect anomalies. This algorithm uses an ensemble of randomized trees to generate anomaly scores. The basic idea behind is to overfit decision tree models. Then AI Studio generates an anomaly score based on how many splits are needed to isolate an instance from the rest of the data points. You also have the option to choose from basic algorithms like IQR and DBSCAN.

Click here to learn the process for detecting the anomalous points of your dataset.