Analysis of Factors Affecting Carbon Emissions based on Gray Correlation Analysis

In order to explore the influencing factors of carbon dioxide emissions, this paper collects data on carbon emissions and several influencing factors that may be related to its changes from 2004 to 2021, considering that too many indicators may make the data redundant and lack of representativeness, using the K-Means clustering algorithm to classify the indicators and defining the names of the corresponding first-level indicators, and analyzing the elbow diagrams for the determination of the model K-value. For the determination of the K value of the model, the optimal number of clusters was analyzed using the elbow diagram, so that the center of each cluster as the first-level indicator data was spliced with the normalized emission data to form a new dataset, and the correlation coefficients between the first-level indicators and the carbon emissions were finally calculated using the grey correlation analysis method. The results show that the ability to mitigate carbon dioxide content has the greatest impact on carbon content, and the correlation coefficient between the two reaches 0.756, which indicates that increasing the green area is very effective in achieving the goal of energy saving and emission reduction.