6.7.4 Detect Anomalies

6.7.4.1 Anomaly Detection Model Types
6.7.4.2 Anomaly Detection Learning Types
6.7.4.3 Anomaly Detection for Logs
6.7.4.4 Prepare Data for an Anomaly Detection Model
6.7.4.5 Train an Anomaly Detection Model
6.7.4.6 Generate Predictions for an Anomaly Detection Model
6.7.4.7 Score an Anomaly Detection Model

Anomaly detection, which is also known as outlier detection, is the machine learning task that finds unusual patterns in data.

MySQL HeatWave AutoML supports unsupervised and semi-supervised anomaly detection. See Anomaly Detection Learning Types to learn more.

The following tasks use datasets generated by OCI GenAI using Meta Llama Models. The anomaly detection use-cases are to find unusual patterns of purchasing behavior for credit card transactions, and to find anomalies in log data.

To generate your own datasets to create machine learning models in MySQL HeatWave, learn how to Generate Text-Based Content.

Note

Datasets were generated using Meta Llama models. Your use of this Llama model is subject to your Oracle agreements and this Llama license agreement: https://downloads.mysql.com/docs/LLAMA_31_8B_INSTRUCT-license.pdf.