Prediction of Coal Prices based on Random Forest and LASSO Regression

Liyuan Li

《International Core Journal of Engineering》 7卷9期 (2021/09) Pp. 67-73

https://doi.org/10.6919/ICJE.202109_7(9).0011

摘要

For coal price influencing factors, firstly, seven factors that may affect coal supply and demand and their data are selected, and a random forest model is used to take the mean value of the contribution made by each influencing factor on each tree to quantitatively investigate the degree of influence of these factors on coal prices and rank them according to their values. A Lasso regression model of coal prices on the influencing factors is then developed through visualisation methods and ten-fold cross-validation to obtain the optimal penalty coefficients based on the minimum mean squared error to forecast monthly coal prices; and through polynomial fitting, forecasts are made for future weekly and daily coal prices. Secondly, using the New Coronavirus outbreak factor as an example, the newly derived equation is modified by introducing dummy variables through multiple regression analysis of the data before the occurrence of the COVID-19, resulting in a modified prediction equation. Finally, in the light of today's social development trends, four policy recommendations are made to safeguard the steady state development of China's coal market.

關鍵字

Coal Prices ； Random Forest ； LASSO Regression ； Polynomial Fit ； Forecasting

參考文獻

An Peng Deng and Yi Xuan Deng. (2013). Conflict of Coal and Electricity Price's Impact on National Economic Development. Advanced Materials Research, 2482pp. 1675-1678.

Xinghua Fan,,Li Wang & Shasha Li.(2016).Predicting chaotic coal prices using a multi-layer perceptron network model. Resources Policy(), doi:10.1016/j.resourpol.2016.08.009.

Ming, Shulei, Song. & Lingyun. (2016). Prediction of China’s coal price during Twelfth Five Year Plan period. Energy Sources, Part B: Economics, Planning, and Policy(6), doi:10.1080/15567249.2012. 689797.

Zhang Wenlong. (2021). Influencing Factors and Projections of Coal Price. IOP Conference Series: Earth and Environmental Science, 632(2), pp. 022016-.

Zhaobin Duan and Duan Zhaobin and Meng Yanjun. (2020). Research on influencing factors of coal price based on improved principal component analysis and grey correlation method. Journal of Physics: Conference Series, 1651(1), pp. 012036-.