基于决策树CART算法的钼金属价格预测方法研究

Research on Method of Mo Price Forecasting Based on Decision Tree CART Algorithm

钼作为一种重要的战略金属,其市场价格受制因素众多。钼金属价格显著的敏感难测性使得构建其预测模型难度较大。本文通过对钼金属价格影响因素的深入分析和现有研究成果的系统梳理,尝试采用决策树CART算法构建钼金属价格预测模型并进行实际价格预测,通过运用统计学原理对结果进行对比分析,其平均绝对误差值为1.61%,预测趋势的准确率可达94.1%,表明该算法不仅简洁直观,而且具有较好的合理性与可靠性。

Mo is an important strategic metal, and its market price is controlled by many factors. As Mo possesses significant price sensitivity and hard to pridict, it is difficult to build its prediction model. This article deeply analyzes the factors which effect the price of Mo, and makes a systematic overview about the existing research results. Trying to use the decision tree CART algorithm to build the prediction model and the Mo price forecasting was made.Via comparing and analyzing the results, by using statistics theory, the mean absolute error is 1.61% and the accuracy of the prediction is 94.1%, which shows that the algorithm is not only concise and intuition, but also has better rationality and reliability.