Optimization Study on Pricing and Replenishment Decision of Vegetable Products Based on LSTM and ARIMA Models

This paper addresses the problem of pricing and replenishment optimization in superstore vegetable merchandise management, using a variety of data processing and analysis methods. Firstly, descriptive statistics and time series analysis are conducted through Excel and MATLAB to reveal the sales volume pattern. Second, correlation coefficient, cluster analysis and SVR regression are applied to explore the cost-plus pricing strategy. Further, the LSTM network is combined to predict the wholesale price and incoming volume to construct the demand price curve. Finally, based on the nonlinear programming model, the optimal replenishment quantity and pricing strategy are optimized considering the constraints of ordering quantity and market demand to achieve the maximum total profit objective. The comprehensive use of ARIMA model and decision variables to establish the optimal single-item shopping strategy provides effective operational decision support for superstores.