基于偏最小二乘支持向量机回归区域物流量预测

Predicting into Regional Logistics Volume Based on Partial Least-squares Support Vector Machines Regression

研究采用偏最小二乘支持向量机回归模型进行区域物流量预测问题.针对普通最小二乘预测所存在的问题和物流系统样本量少的具体状况,提出偏最小二乘支持向量机回归区域物流量预测方法,采用主成分分析法提取影响物流量因素的新综合变量,建立以新综合变量为输入,物流量为输出的支持向量机回归非线性预测模型,在廊坊市物流量预测中进行仿真试验,证明了该方法的可行性与正确性.

The prediction problem of regional logistics volume was studied by using partial least-squares support vector machines regression model. Bases on analyzing the exiting problem about normal least-squares regression and the concrete situation of few samples in regional logistics system, the partial least-squares support vector machines regression method of regional logistics volume was brought forward. The new synthetic variables were extracted about influencing logistics factors by principal component analysis method. The nonlinear predicting model by using support vector machines regression was established which the new synthetic variables were used as the input variables of the model and the logistics volume was used as the output variable. The simulation experiment had been done in the logistics system of Langfang City. The result shows that the method is correct and feasible.