Unsupervised methods to identify unusual or rare events in the data. In trading, anomalies could be flash crashes, erroneous price spikes, or regime shifts. Clustering or statistical models can flag points that deviate significantly from typical patterns. For example, an unsupervised model could monitor tick data and signal an anomaly when the current price jump far exceeds historical volatility. Detecting anomalies helps a forex robot avoid executing trades during erratic market conditions or alert on potential data errors.

Back to Robots Glossary Index