随着铁路部门信息化的建设,海量的数据积累使得采用数据挖掘技术对铁路货运量进行预测十分必要。通过系统分析、数据预处理、数据挖掘及知识提取,提出了预测铁路货运量的三种算法:线性回归、BP神经网络及支持向量回归机,并通过实例验证比较了算法的有效性。 With the construction of railway information department, massive data accumulation makes it necessary to use data mining technology to forecast the freight volume of railway. Through the system analysis, data preprocessing, data mining and knowledge extraction, three algorithms for forecasting railway freight volume are put forward: linear regression, BP neural network and support vector regression machine. The effectiveness of the algorithm is verified by an example.