ECONOMETRIC MODELING VS. MACHINE LEARNING: A COMPARATIVE ANALYSIS OF PREDICTIVE ACCURACY AND PARAMETER INTERPRETABILITY IN ECONOMIC FORECASTING

This article examines the theoretical and methodological aspects and practical applications of machine learning algorithms in econometric modeling. It provides a comparative analysis of the characteristics, advantages, and limitations of classical econometric models and modern machine learning algorithms. The study compares linear regression and time series-based econometric approaches with machine learning algorithms such as Random Forest, Gradient Boosting, and XGBoost. The findings provide a methodological basis for selecting an appropriate modeling approach based on data characteristics and research objectives in the analysis of economic indicators.