This study employs machine-learning algorithms (ML), specifically Random Forest (RF) and Gradient Boosting (GB), to assess the impact of various factors, including Gross Domestic Product (GDP) growth, urbanization, and energy consumption, on carbon dioxide emissions (CO2). The research underscores the RF algorithm's superior accuracy in determining independent variables' influence on CO2 emissions compared to GB. Furthermore, the study reveals that natural gas is the most significant contributor to CO2 emissions in Egypt, accounting for 49.7% of the total, followed closely by oil at 46.7%. The effect of other variables on CO2 emissions is relatively minimal. The findings also establish a strong positive correlation between the consumption of natural gas, oil, and coal and CO2 emissions in Egypt. Additionally, a negative relationship is observed between GDP growth, suggesting a positive trend in environmentally friendly economic expansion and urbanization on CO2 emissions in Egypt. This unique scenario, where urban expansion appears to have an inverse relationship with CO2 emissions, sets Egypt apart from many other countries and signifies a favorable environmental outcome.
Read full abstract