节假日电商打折中物流量突变预测方法研究  

Study on Forecasting of Sudden Changes in Logistics Volume during E-commerce Holiday Discounts

作　　者：牛三勇[1] 

机构地区：[1]宁夏大学经济管理学院,宁夏银川750021

出　　处：《物流技术》2014年第3期160-162,共3页Logistics Technology

摘　　要：针对节假日电商打折中物流量的变化呈现出较大的突变性、预测效果不好的问题,提出一种基于多元逐步回归的节假日电商打折中物流量突变的预测方法。首先采用多元线性回归分析法,选择节假日电商打折中物流需求量的影响因子,然后通过遗传算法对最小二乘支持向量机的参数进行优化,最后建立节假日电商打折中物流需求量与影响因子之间的非线性关系模型,完成物流需求的准确预测。系统测试结果表明,相对于参比模型,MLR-LSSVM提高了物流需求量的预测精度,能够准确刻画节假日电商打折中物流量的变化趋势。In this paper, in view of the characteristics of the changes in logistics volume during e-commerce holiday discounts, we proposed a method for their forecasting based on multiple stepwise regression. First, using multiple linear regression analysis, we selected the influence factors of the logistics demand volume in e-commerce holiday discounts, then optimized the parameters of the least square SVM using genetic algorithm and at the end built the nonlinear relational model of the logistics volume during the abovementioned discounts and the influence factors. At last, through a system testing, we proved the advantage of the model in forecasting the sudden changes in logistics volume during the discounts.

关 键 词：物流需求 节假日 最小二乘支持向量机 电商打折 预测方法 

分 类 号：F252.21[经济管理—国民经济] F713.365.1