基于Bayes-BP算法的跨境电商平台采购量预测  

CROSS-BORDER E-COMMERCE PLATFORM PURCHASE VOLUME FORECAST BASED ON BAYES-BP ALGORITHM

作　　者：孙桐 徐斌 贾航 Sun Tong;Xu Bin;Jia Hang(School of Shipping Economics and Management,Dalian Maritime University,Liaoning Provincial Key Laboratory of Logistics and Shipping Management System Engineering,Dalian 116026,Liaoning,China)

机构地区：[1]大连海事大学航运经济与管理学院辽宁省物流航运管理系统工程重点实验室,辽宁大连116026

出　　处：《计算机应用与软件》2021年第12期91-96,共6页Computer Applications and Software

摘　　要：传统预测采购量采用时间序列,预测模型较为单一。而平台采购量与消费者的多种消费行为有关。故采用Granger因果检测,得到影响商品销量相关性较高的几种因素,如商品评论量、商品收藏量、是否有税费补贴等;利用Bayes-BP算法建立模型,优化传统的BP算法,得到商品采购量。将该方法应用于大连某跨境电商平台商品采购量预测,结果表明该方法的平均误差率在5.9%,具有较高实际应用价值。The traditional forecasting purchase volume uses time series,and the forecasting model is relatively simple.The amount of platform purchases is related to consumers various consumption behaviors.This paper used Granger causality testing to obtain several factors that have a high correlation with product sales,such as the number of product reviews,the number of product collections,and the existence of tax subsidies.This paper used Bayes-BP algorithm to build a model,optimized the traditional BP algorithm,and got the commodity purchase volume.This method was applied to forecast the purchase volume of a cross-border e-commerce platform in Dalian.The results show that the average error rate of this method is 5.9%,which has high practical application value.

关 键 词：采购预测 Bayes-BP算法 Granger因果检测 跨境电商平台 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] U461.2[机械工程—车辆工程]