Clustering Model-based Analysis of Sales in Vegetable Categories for Supermarkets

The development of effective vegetable sales strategies is crucial for enhancing supermarket profitability, and these strategies are heavily reliant on a nuanced understanding of the sales volume distribution patterns across different vegetable categories. However, most existing studies fail to thoroughly analyze these distribution patterns, often neglecting factors such as sales volume magnitude and fluctuations. To address this gap, we first propose a comprehensive framework based on a K-means++ clustering model and Spearman correlation coefficient to analyze the sales data. Firstly, we normalize the data to eliminate the impact of differing dimensions across vegetable categories. Next, we analyze the sales data from both quarterly and daily perspectives. Data visualization is performed utilizing line charts, followed by a descriptive analysis of the sales volume distribution considering both magnitude and fluctuation. Furthermore, we utilize a K-means++ clustering model and the Spearman correlation coefficient to further explore the distribution patterns and interrelationships among different vegetable categories. The results of the experiments reveal the daily and quarterly sales distribution patterns of various vegetable categories, providing valuable insights to support the development of effective vegetable sales strategies for supermarkets.