A Study of Fresh Commodities Sales Analysis Based on Time Series Modeling - Taking Vegetable Category Commodities as an Example

Abstract

This paper serves as a decision-making aid for pricing and replenishing vegetables in fresh produce supermarkets. Initially, the sales data of various vegetable categories are thoroughly examined for trends and seasonal patterns, utilizing linear regression models that account for temporal trends and quarterly dummy variables. The analysis reveals that the four primary categories of flower vegetables, solanaceous vegetables, capsicum vegetables, and edible fungi exhibit distinct trends in their time-series data, while all six categories exhibit notable seasonal patterns. Furthermore, a correlation analysis is conducted on the sales of vegetable categories, considering daily, monthly, and quarterly timescales. The findings reveal a pronounced positive correlation among the sales of individual vegetable categories on a daily basis. Additionally, the seasonal influence on leaf vegetables and edible fungi is observed to be largely similar. Lastly, a K-Means clustering technique is employed to categorize vegetable items based on their sales data. The results indicate that the sales of the ten vegetable items grouped along with Yunnan lettuce and Chinese cabbage are significantly higher, providing valuable insights for inventory management and pricing strategies.