Optimal Strategy Research on Vegetable Replenishment and Pricing Based on Machine Learning Regression

Abstract

The purpose of this study is to investigate the relationship between sales volume and price in the process of selling vegetables in supermarkets, and in turn, to obtain the optimal replenishment and pricing strategy for the next 7 days in supermarkets under profit maximization. By comparing six machine learning regression algorithms, namely Decision Tree, Random Forest, Plain Bayes, Support Vector Machine, XGBoost and CatBoost, the CatBoost model was found to have the best fit. After model tuning and optimization, the relationship between sales volume and price of each category of vegetables in different quarters was obtained. Finally, the objective revenue function was constructed by considering factors such as pricing, sales volume and profit margin, and the particle swarm optimization algorithm was used to determine the optimal pricing and replenishment strategies corresponding to the maximum revenue. The in-depth analysis and modeling of vegetable sales data provide important references for superstores to formulate reasonable pricing and sales plans in different time periods to enhance operational efficiency and market competitiveness.