Prediction and Analysis of Carbon Emissions of Chinese Provinces

Abstract

The paper makes forecasts of the carbon emissions of Chinese provinces to provide references for the government's emission reduction policy. The methods to analyze carbon emissions include K-means clustering and logistic regression analysis. The data about carbon emissions for each province in China during 1995~2011were collected from the statistical database of China's economic and social development (CNKI). The results show that the forecast value of carbon emissions in 28 provinces and regions in China in 2011~2020 is 488,232~6,618,514 million tons. The conclusion of the paper is that the accuracy of Logistic model is influenced by the value of K. As long as the value of K is determined in a reasonable range, it can give a relatively high fitting degree and low prediction error.