Akademik Veri Yönetim Sistemi
National Accounting Review, cilt.7, sa.4, ss.501-521, 2025 (ESCI, Scopus)
This study analyzed the determinants of gross domestic product (GDP) for Germany and Poland using both linear econometric models and nonlinear machine learning models (decision trees, random forests, XGBoost) on data from 1991 to 2023. By comparing the model outcomes for Germany and Poland, we identified structural differences and uncovered key predictors of economic growth, measured by gross domestic product, over 33 years. Empirical results showed that nonlinear models significantly outperformed linear ones, with XGBoost achieving the best results in Germany, while the decision tree performed best in Poland. We also conducted feature importance analysis to reveal key factors. For Germany, factors such as life expectancy, net migration, and foreign direct investment were the strongest predictors of GDP. In Poland, production volume, life expectancy, urban population, internet usage, foreign direct investment, and unemployment rate emerged as the key drivers of GDP. Our insights highlight the need for specific economic modeling strategies and show how different development paths shape national growth dynamics.