A machine learning paradigm where models are trained on labeled data, learning to map input features (e.g. historical forex indicators) to known outputs (e.g. price targets or buy/sell labels). Supervised methods (such as regression or classification) are widely used for predictive tasks in trading, e.g. forecasting currency prices or classifying market regimes. In practice, data is split into training and validation sets (e.g. X_train, y_train), and one calls routines like model.fit(X_train, y_train) to train models. Loss functions (see below) quantify prediction errors during this process. Models learned via supervised learning can then make predictions (model.predict()) or probabilistic forecasts (model.predict_proba()) on new forex data.

Back to Robots Glossary Index