Carbon Dioxide Emission by Cars: A Performance Analysis of Diverse Machine Learning Regressors and Interpretation using Explainable AI

This study introduces a creative paradigm for Emission of Carbon by integrating machine learning and explainable artificial intelligence (XAI) techniques. A variety of regression methodologies, including Ridge, Linear, Support Vector, and Lasso regressions, were used to select specific machine learning algorithms. Additionally, the regression and regressor ensemble learning framework were employed. Through a comparative analysis, boosting algorithms, especially XG Boost, demonstrate remarkable error. This enhanced the final XG Boost model’s error to a magnificent 9.195e-07. The incorporation of XAI methods, such as SHAP values and LIME, provides an intricate understanding of feature importance, feature contribution and local explanation, elevating model interpretability. This research’s significance lies in its potential to transform carbon emission screening, offering low error precision, resilience, and transparency. The combination of machine learning regressors and XAI techniques contributes to advanced screening tools, fostering trust among ecologist and researchers.