Predicting Real GDP Growth Using Machine Learning Models: Random Forest vs. XGBoost

Objective:

• To predict cross-country real GDP growth rates based on macroeconomic indicators.

• To explore the effectiveness of machine learning approaches as an alternative to traditional economic models.
  
  

Data

Data Sources:

• World Bank World Development Indicators

Time Period and Coverage:

• Years: 1991–2023

• Countries: Over 30 countries (adjust depending on the actual number)
  
  

Key Variables:

• Inflation rate

• Unemployment rate

• Government expenditure-to-GDP ratio

• Trade openness ((exports + imports)/GDP)

• Private sector credit growth

Models and Methodology

• Random Forest (ntree = 100)

• XGBoost (nrounds = 50, objective = "reg:squarederror")

•  Evaluation Metric: Mean Squared Error (MSE)

Results

Model Performance Comparison (MSE):

Random Forest: 5.1183

XGBoost: 5.5995

•  Random Forest achieved slightly better prediction accuracy than XGBoost.

Actual vs Predicted GDP Growth Plots:

• Random Forest:
  Predicted values aligned closely with the 45-degree ideal line, indicating strong predictive accuracy.