The process of transforming raw data into meaningful input features for a model. It involves selecting, creating or transforming variables so that a learning algorithm can use them effectively. In forex AI, feature engineering might include computing technical indicators (moving averages, RSI), combining multiple price/time features, or deriving sentiment scores from news. Good features allow a machine-learning model to learn more accurate trading signals. For example, turning tick data into trend/momentum features can greatly improve a model’s ability to predict price moves.

Back to Robots Glossary Index