Socioeconomic drivers and environmental pressure on renewable energy in Azerbaijan: Machine learning evidence

  • 8 Views
  • 3 Downloads

Creative Commons License DMCA.com Protection Status
This work is licensed under a Creative Commons Attribution 4.0 International License

Type of the article: Research Article

Abstract
This study explores potential factors associated with renewable energy consumption in Azerbaijan by examining the relationships among economic growth, human development, per capita CO₂ emissions, and urban population growth. The analysis applies an exploratory machine learning framework based on the XGBoost algorithm combined with SHapley Additive exPlanations (SHAP) to examine possible nonlinear associations and evaluate the relative contribution of selected socioeconomic and environmental variables. Correlation analysis indicates a negative relationship between renewable energy consumption and CO₂ emissions (−0.55), while human development and economic growth exhibit generally positive associations with renewable energy use, suggesting a possible role of socioeconomic development in the energy transition process. The exploratory model produced performance indicators of R² = 0.81, RMSE = 1.18, and MAE = 1.04, suggesting that the model captures variation within the available sample; however, given the limited dataset, these results should not be interpreted as evidence of strong or robust predictive performance. SHAP-based interpretation suggests that human development and environmental pressure may represent relatively important variables within the model framework, while the effects of economic growth and urbanization appear comparatively moderate and context-dependent. Overall, the study provides preliminary exploratory evidence regarding possible nonlinear patterns linking socioeconomic development, environmental conditions, and renewable energy consumption in Azerbaijan, offering directions for future research using larger datasets and complementary empirical approaches.

view full abstract hide full abstract
    • Figure 1. Time-series dynamics of selected indicators
    • Figure 2. Correlation of renewable energy and socioeconomic variables
    • Figure 3. XGBoost feature importance of renewable energy factors
    • Figure 4. SHAP feature importance of renewable energy predictors
    • Figure 5. SHAP value distribution of features
    • Table 1. Descriptive overview of selected variables
    • Table 2. XGBoost model performance metrics
    • Table 3. Feature importance (XGBoost gain)
    • Conceptualization
      Anar Eminov, Ramil Hasanov, Jeyhun Mahmudov
    • Formal Analysis
      Anar Eminov, Ramil Hasanov, Jeyhun Mahmudov, Abbas Musayev, Mekhdi Bagirov
    • Investigation
      Anar Eminov, Ramil Hasanov, Jeyhun Mahmudov, Abbas Musayev, Mekhdi Bagirov
    • Methodology
      Anar Eminov, Ramil Hasanov
    • Project administration
      Anar Eminov, Ramil Hasanov, Jeyhun Mahmudov
    • Supervision
      Anar Eminov, Abbas Musayev, Mekhdi Bagirov
    • Writing – original draft
      Anar Eminov, Ramil Hasanov
    • Data curation
      Ramil Hasanov
    • Software
      Ramil Hasanov
    • Validation
      Ramil Hasanov, Jeyhun Mahmudov, Abbas Musayev, Mekhdi Bagirov
    • Writing – review & editing
      Jeyhun Mahmudov, Abbas Musayev, Mekhdi Bagirov