A supervised learning task where the model predicts a continuous numeric output. In forex trading, regression models (e.g. linear regression, polynomial regression) forecast continuous values like future exchange rates or returns. For example, a regression-based forex robot might learn a function mapping recent price history to next-day price. Common regression algorithms (linear regression, support vector regression, decision-tree regression) attempt to fit a function to historical price data; their performance is often evaluated with metrics like mean squared error. Regression is crucial when the goal is precise price-level forecasting rather than class labels.

Back to Robots Glossary Index