Research on Forecasting Vegetable Sales and Prices Based on Time Series Models

Aiming at the characteristics of fresh products that are easy to deteriorate in room temperature environment, superstores need to develop reasonable pricing strategies to improve profitability. This paper analyzes the sales data of vegetables in a superstore, uses Pearson correlation analysis and K-prototype cluster analysis to find the connection between the sales of different categories or different single vegetables, and uses the ARIMR time series forecasting model to forecast the sales and profits of different categories of vegetables. The results show that the ARIMR time-series forecasting model has good accuracy in predicting sales and profits and can provide analysis and support for pricing and replenishment decisions in supermarkets.