Predicting carbon dioxide emissions in Vietnam based on factor groups: economic, social, investment, and industry using a deep learning model

Over the past two decades, Vietnam has faced severe consequences from climate change, including floods, droughts, and increasing environmental pollution. One of the leading causes is rising CO2 emissions, primarily from industrial production activities and energy consumption. This study aims to develop a deep learning model for predicting CO2 accuracy based on the factors influencing the increase in carbon emissions in Vietnam. These factors were identified by analyzing the correlation relationships between them and CO2 emissions. This study also proposes a data transformation method and applies it to the prediction model to improve the accuracy of the prediction result. The results show that with an RMSE error of 0.025 and an R-squared of 0.97, the model demonstrates the performance in forecasting CO2 emissions, significantly surpassing traditional methods like K-nearest neighbors (KNN), support vector regression (SVR), and Naive Bayes, and reflecting its effectiveness in predicting future carbon emission trends. This method not only provides detailed insights into the relationships between influencing factors but also predicts CO2 emissions to 20230, which helps policymakers in developing appropriate carbon reduction strategies.