Abstract: Accurate prediction of coal price can improve the scientific sales decision on coal . An interval combination prediction model of coal price based on improved weighted clustering is proposed to improve the accuracy of coal price prediction. This paper analyzes the factors affecting coal production and production from the aspects of coal price and inventory. According to the volatility characteristics of coal price data, the wavelet transform function is defined, and the preprocessing of coal price data is completed by eliminating the noise of coal price data. On the basis of introducing the concept of interval number separation degree in multi-attribute decision-making, the improved weighted clustering method is used to determine the prediction weight of coal price interval combination. By calculating the weighting coefficient of coal price interval combination prediction, the coal price interval combination prediction model is established to realize the prediction of coal price. The simulation results show that the root mean square error and average absolute error can be controlled between 0.1 ~ 0.3 and 0.2 respectively, which greatly improves the prediction accuracy of coal price.

Key words: improved weighted clustering, interval type combination, influencing factors, coal price, prediction model, production costs