Urban transport emission prediction analysis through machine learning and deep learning techniques

Abstract

About 6.6 million people die every year from air pollution diseases globally. Transportation industry is considered one of the leading contributors in air pollution. This research utilizes deep learning and machine learning techniques to predict China’s transport-related CO₂ emissions and energy needs by utilizing variables like population, car kilometers, year and GDP per capita. The outcomes have been analyzed using six analytical measures: determination coefficient, RMSE, relative RMSE, mean absolute percentage error, mean bias error and mean absolute bias error. Findings indicate that yearly increase in transport-related CO₂ emissions in China will be 3.66%, and transport energy consumption will increase by 3.8%. Energy consumption and transport CO₂ emissions are projected to rise by roughly 3.5 times by 2050 as compared to current levels. Therefore, government should re-evaluate its energy investment plans for the future and institute new rules, and standards regarding transport-related energy consumption and pollution reduction.