Clustering Data Penjualan Kopi Sidikalang Menggunakan Metode K-Means

This study aims to analyze the sales patterns of Sidikalang coffee using the K-Means clustering method to optimize stock management and marketing strategies. Three months of sales data were analyzed based on the attributes of sales volume, price, and season. The results showed that the data could be grouped into four clusters with distinct characteristics: low sales (52 data points), medium sales (50 data points), high sales (21 data points), and high stock with sporadic sales (77 data points). Evaluation using the Davies-Bouldin Index (DBI) yielded a score of 0.52, indicating good clustering quality. These findings provide valuable insights for business actors in developing more targeted marketing strategies and efficient stock management. This research also contributes to the literature on the application of machine learning in analyzing local product sales, particularly Sidikalang coffee.