基于ARIMA和BP神经网络的供应链需求预测模型及其对比分析

Supply Chain Demand Forecasting Model Based on ARIMA and BP Neural Network and Its Comparative Analysis

摘要: 提升整个供应链效益首要解决的任务是牛鞭效应，而提高预测精准度可以有效的抑制牛鞭效应。基于此在深入分析ARIMA模型和BP神经网络特性的基础上，以某商城办公用品的数据为例对需求量进行预测，选取绝对百分比误差MAPE、均方根误差RMSE和平均绝对误差MAE作为模型检验指标，比较相对误差衡量两种预测模型。构建有较高预测精度的模型，为企业应对市场需求变化提供重要的理论依据。

Abstract: The primary task to be solved to improve the efficiency of the entire supply chain is the bullwhip effect, and improving the accuracy of prediction can effectively suppress the bullwhip effect. Therefore, based on the in-depth analysis of the ARIMA model and the BP neural network characteristics, this paper takes the data of a certain mall office supplies as an example to predict the demand, and selects the absolute percentage error MAPE, the root mean square error RMSE and the average absolute error MAE as the model test indexes that compare the relative errors to measure the two forecasting models. Constructing a model with higher prediction accuracy provides an important theoretical basis for enterprises to respond to changes in market demand.