基于SVM的区域物流需求建模与预测仿真——以浙江省为例  

Modeling and Forecasting Simulation of Regional Logistics Demand Based on SVM-Take Zhejiang Province as an Example

作　　者：梁毅 徐超飞 LIANG Yi;XU Chao-fei(School of Management,Hebei GEO University,Shijiazhuang 050031,China)

机构地区：[1]河北地质大学管理学院,河北石家庄050031

出　　处：《物流研究》2024年第3期54-60,共7页Logistics Research

基　　金：教育部人文社会科学研究青年基金项目“面向区域低碳能源统筹规划的电能替代实施路径研究”(21YJC630072)。

摘　　要：为了提升区域物流需求预测的准确度,提出了一种基于支持向量机的物流需求预测方法。首先,通过主成分分析法对区域物流需求影响指标进行筛选,然后输入样本数据进行学习,最终建立区域物流需求与影响指标之间的非线性模型。基于浙江省2002—2021年的物流需求进行仿真分析,结果显示,相比于BP神经网络,SVM在区域物流需求预测方面有更高的预测精度,具有广泛的应用前景。In order to improve the accuracy of regional logistics demand forecasting,a logistics demand forecasting method based on support vector machine(SVM)is proposed.Firstly,the dimensionality reduction of the influencing indicators of regional logistics demand is carried out by principal component analysis(PCA)method,and then the samples are input to SVM for learning,and finally a nonlinear model between regional logistics demand and influencing indicators are established.The test of logistics demand in Zhejiang Province from 2002 to 2021 shows that SVM has higher prediction accuracy in regional logistics demand forecasting than the BP neural network model,and has a wide range of application prospects.

关 键 词：区域需求预测 SVM 影响因素 BP神经网络 

分 类 号：F259.27[经济管理—国民经济]