Research on Iron Ore Price Prediction Based on AdaBoost-SVR

Abstract

This study aims to use the support vector regression (SVR) theory, according to the nonlinear characteristics of iron ore price series fluctuation, based on the 5000 daily transaction data of iron ore in Dalian Commodity Exchange as the research object, the Adaboost -SVR iron ore price prediction model optimized by the novel BAT algorithm (NBA) is established. The model takes the maximum, minimum, closing price and trading volume of the daily transaction data as input parameters and the closing price of the next trading day as output parameters. The prediction results of the research model are compared and analyzed. The results show that the prediction value of the research model is closer to the real value, and the mean relative error (MRE) and root mean square error (RMSE) of the research model are 0.006 and 20.19, respectively, which are better than the prediction results of the traditional support vector regression model. The research model provides technical support and decision-making basis for the market monitoring and early warning of iron ore, and has advantages in accuracy compared with traditional forecasting methods.