Grey relational analysis, principal component analysis and forecasting of carbon emissions based on long short-term memory in China

Abstract

With the development of China's economy, the use of fossil energy has become more and more, resulting in increasing carbon emissions. CO2 emissions have caused global warming, threatening humans and creatures on Earth. In order to effectively suppress the growth of carbon emissions, it is necessary to analyze the influencing factors of carbon emissions and apply them to predict carbon emissions. This paper presents sixteen potential influencing factors and uses grey relational analysis to identify the factors that have a strong correlation with carbon emissions. The principal component analysis (PCA) is used to extract the four principal components, which reduce the redundancy of the input data. The long short-term memory (LSTM) method is established to predict carbon emissions in China. We use back propagation neural network (BPNN) and Gaussian process regression (GPR) to compare LSTM method. The simulation results show that the prediction accuracy of carbon emissions based on LSTM is better than that of BPNN and GPR, indicating the effectiveness of PCA and LSTM in prediction of carbon emissions. Finally, this paper provides the theoretical basis for China to reduce carbon emissions by studying prediction of carbon emissions.