Techniques that combine multiple models or trading strategies to improve overall performance. An ensemble might average or “vote” on signals from several algorithms, reducing reliance on any single model. In machine learning, ensemble methods (like random forests, boosting, or a portfolio of diverse alphas) are proven to yield better predictive accuracy than individual models. In practice, a Forex robot might use an ensemble of mean-reversion and momentum models, for example, to smooth returns and hedge model-specific errors.

Back to Robots Glossary Index