Automatic Pricing and Replenishment Decision of Vegetable Commodities: A Multi-model Perspective

In fresh food supermarkets, the shelf life of vegetable products is relatively short. In order to make reliable and reasonable replenishment and pricing decisions, and achieve maximum sales revenue and profit, this article first studies the Pearson correlation coefficient model to study the relationship between various categories of vegetables and the sales volume of individual products; Then, a time series model is established from a vertical perspective combined with ARIMA, and a random forest model is established from a horizontal perspective. Then, the final strategy for predicting the daily replenishment volume and pricing strategy for the next week is obtained by integrating the two models; Finally, considering the limited sales space of vegetable supermarkets, we optimized the machine learning model using gradient descent and proposed a daily replenishment and pricing strategy to maximize profits.