摘要
为提高大数据背景下跨境电商皮革服装销量预测的精度,本文提出一种基于BP神经网络模型的预测方法。方法以BP神经网络为基础预测模型,通过采用遗传算法优化BP神经网络的初始权值阈值与网络结构,克服BP神经网络收敛速度慢和容易陷入局部最优的问题。采用遗传算法优化的BP神经网络构建预测模型,用于跨境电商皮革服装销量预测,提高了皮革服装销量预测的精度。仿真结果表明,相较于改进前的标准BP神经网络与ARMA模型、Grey Model模型和改进Grey Model模型,所提的遗传算法改进BP神经网络模型的平均绝对误差、平方和误差、均方误差更小,更能反映真实的跨境电商皮革服装销售趋势,具有一定的实际应用价值。
A prediction method based on BP neural network model was proposed to improve the accuracy of cross-border e-commerce leather clothing sales forecast under the background of big data.Methods based on the BP neu⁃ral network,the genetic algorithm was used to optimize the initial weight threshold and network structure of the BP neural network,so as to overcome the problems of slow convergence speed of the BP neural network and easy to fall into local optimum.BP neural network optimized by genetic algorithm was used to build a prediction model,which was used to forecast the sales volume of cross-border e-commerce leather clothing,improving the accuracy of leath⁃er clothing sales forecast.The simulation results show that compared with the standard BP neural network and ARMA model,Grey Model and Grey Model before improvement,the average absolute error,sum of squares error and mean square error of the proposed genetic algorithm improved BP neural network model are smaller,which can better re⁃flect the sales trend of real cross-border e-commerce leather clothing,and has certain practical application value.
作者
李振宏
LI Zhen-hong(College of Business,Nanning University,Nanning 530200,China)
出处
《中国皮革》
CAS
2023年第6期104-109,共6页
China Leather
基金
2022年度广西高校中青年教师科研基础能力提升项目(2022KY1769)。
关键词
大数据
服装销量预测
BP神经网络
遗传算法
big data
clothing sales forecast
BP neural network
genetic algorithm
