Gross Domestic Product (GDP) is one of the critical indicators of an economy. This study aims to predict the GDP of the United Kingdom using vital macroeconomic variables from 1990 to 2018 as predictors, which include energy prices, unemployment rate, Real Effective Exchange Rate (REER) inflation and net migration. Several machine learning models, namely Support Vector Regression (SVR), Random Forest (RF) and Gradient Boosting Machines (GBM), were implemented, analysed and compared. The models were trained on both scaled and unscaled data, with hyperparameter tuning applied to optimise performance. The models’ performances and accuracy were analysed by employing evaluation metrics Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). As per the findings, after hyperparameter tuning, the SVR model performed best in GDP prediction, followed by GBM. The results of this study underscore the critical role of macroeconomic variables in GDP prediction while highlighting the potential of machine learning models to produce valuable and reliable insight into economic forecasting.
Read full abstract