Research on Chinese Coal Forecast Analysis Based on RF-GINI Algorithm

Under the current situation of Chinese economic growth, population development and energy revolution, the influencing factors of coal consumption demand have received extensive attention. Based on the main influencing factors in the three directions of economy, population and energy, the data of relevant indicators from 2001 to 2019 are retrieved and collected. Some indicators are eliminated by the kendall correlation coefficient. After PCA (principal component analysis) dimension reduction processing, RF (random forest) model training is established to iterate the data, and the importance of the main influencing indicators of GINI coefficient research is analyzed. The results show that the influencing factors of Chinese coal consumption are ranked from the highest to the lowest: rural population, total crude oil consumption, urban population, GDP, total energy consumption and national fiscal and tax revenue. Provide a reference for the decision-making of relevant departments.