基于季节性ARIMA模型的新零售精准预测

Accurate Forecast of New Retail Based on Seasonal ARIMA Model

本文旨在建立新零售的精准预测模型,首先通过对新零售目标商品的主要指标数据进行数据预处理,接着建立Pearson相关系数模型,使用Python分析得到热力图,确定销售量具有较好的预测性并存在自相关性,将其作为本文预测模型的重要决策变量,然后建立差分整合移动平均自回归模型(Autoregressive Integrated Moving Average model,简称ARIMA),同时考虑商品由于季节所造成的影响,并优化成季节性ARIMA预测模型,最后使用平均绝对百分比误差(MAPE)评估模型,得到预测误差百分比均值为19.33%。本文模型预测误差小,对新零售商品的预测具有指导意义。

The purpose of this paper is to establish an accurate prediction model of new retail. Firstly, the main index data of new retail target goods are preprocessed, and then Pearson correlation coefficient model is established. The thermodynamic diagram is obtained by Python analysis. The sales volume is confirmed to have good predictability and existing autocorrelation. Then the Autoregressive Integrated Moving Average model (ARIMA) is established. At the same time, the seasonal ARIMA prediction model is optimized considering the same influence of the season. Finally, the average absolute percentage error (MAPE) evaluation model is used to obtain the average forecast error percentage of 19.33%. The prediction error of the model in this paper is small, which has guiding significance for the prediction of commodities.