由于煤炭安全投入影响因素较多,所具有的非线性、复杂性与内在随机性,传统的方法难以准确模拟其变化状态并预测未来需求。神经网络是一种较好解决此类问题的方法,但是本身存在诸多缺陷,而这些缺陷支持向量机(Support Vector Machine,SVM)可以克服。本文基于安全投入预测的非线性,利用统计学习理论的支持向量机,建立基于支持向量机的预测模型。 Due to the nonlinear, complex and inherent randomness of many factors affecting the coal safety investment, it is difficult for the traditional methods to accurately simulate the changing conditions and predict the future demand. Neural network is a better way to solve such problems, but there are many shortcomings in itself, and these defects can be overcome by Support Vector Machine (SVM). Based on the non-linearity of safety investment forecasting and the support vector machine based on statistical learning theory, this paper establishes a forecasting model based on support vector machine.