基于遗传神经网络下新零售目标产品的销量预测研究

Research on Sales Forecast of New Retail Target Products Based on Genetic Neural Network

产品的短期销售需求预测对减少零售企业的库存管理压力,节约运营成本具有重要意义。利用某零售企业生产的N款产品在华东区内的相关数据,运用统计回归方程模型得出目标SKC(单款单色产品)在节假日的销售量。基于此,使用遗传-神经网络算法对10个目标小类在12周内对每周的周销量进行预测,求出各周的MAPE,并预测目标小类内所有SKC在12周内每周的销量。对于目标小类内的SKC,根据销售量变化趋势与所属目标小类变化趋势进行对比,对于差异较大的SKC,用折线图表现其与所属目标小类变化趋势的差异,实现对目标产品短期销量的精准预测。

The short-term products sales’demand forecast is of great significance to reduce the inventory management pressure of retail enterprises and save operating costs.Based on the relevant data of N products produced by a retail enterprise in East China,the sales volume of targeted SKC(stock keeping color)in holidays is obtained by using statistical regression equation model.Based on this,this paper uses genetic neural network algorithm to predict the weekly sales of 10 targeted subcategories in 12 weeks,calculates the weekly MAPE of each week,and forecasts the weekly sales of all SKCS in the target subcategories in 12 weeks.For the SKC within the targeted subcategories,according to the changing trend of its sales volume and that of its targeted subcategories,a comparison is made.For the SKC with large difference,a line chart is used to show the differences between the changing trend of its sales volume and that of its targeted subcategories,so as to realize the accurate prediction of the short-term sales volume of the targeted products.