Machine Learning Approaches for Modeling Economic Growth and Sectoral Performance Dynamics

Abstract

Economic growth and sectoral performance are shaped by complex interactions among macroeconomic indicators, technological investments, labor productivity, and global market shocks. Traditional econometric models often struggle to capture nonlinear relationships and high-dimensional data patterns. This study explores machine learning approaches—including Random Forest, Gradient Boosting, Support Vector Regression, and Long Short-Term Memory networks—to model and predict economic growth and sectoral dynamics across manufacturing, services, agriculture, and energy sectors. The research integrates historical macroeconomic data such as GDP, inflation, interest rates, FDI, export volumes, workforce productivity, and digital adoption indices. Results demonstrate that machine learning techniques outperform classical regression models in forecasting accuracy, pattern detection, and sensitivity to structural economic shifts. Random Forest and Gradient Boosting perform exceptionally well in variable importance estimation, while LSTM captures long-term temporal dependencies in sectoral growth trajectories. The findings highlight the potential of machine learning to support policymakers, financial analysts, and development planners in designing informed strategies, evaluating investment priorities, and mitigating economic vulnerability. Overall, this study demonstrates the transformative role of machine learning in understanding economic evolution, supporting high-quality predictions, and guiding evidence-based decision-making for sustainable economic development.