Comparative Analysis of Predictive Models for Carbon Emission in Major Countries: A Focus on Linear Regression and Random Forest

Mainak Mitra, Soumit Roy

Abstract: This study employs advanced predictive modeling techniques to examine carbon emissions across significant countries. Utilizing a comprehensive dataset from 1757 to 2017, it delves into the emission patterns of countries with the highest and lowest emissions. The study compares the efficacy of Linear Regression and Random Forest Regression models in predicting carbon emissions for Bangladesh, China, India, and the United States. The results, favoring the Random Forest model based on reduced Mean Squared Error, also project future emissions for these countries over the next 50 years. This research contributes to the discourse on sustainable environmental practices and policy-making by providing a solid foundation for understanding and forecasting carbon emission dynamics.

Keywords: Carbon Emission, Predictive Modeling, Linear Regression, Random Forest Regression, Sustainability