Penerapan Algortima K-Means Clustering untuk Optimalisasi Persediaan Liquid Vape Berdasarkan Data Penjualan

Liquid vape is a liquid in an electronic cigarette (vape) device that contains a mixture of Propylene Glycol (PG), Vegetable Glycerin (VG), flavorings, and contains nicotine. As the use of vapes increases as an alternative to conventional cigarettes, efficient stock management becomes a challenge for vape shops to be able to meet customer needs without experiencing excess or shortage of inventory. Good stock management in a retail business is very important to maintain a balance between demand and product availability. This research aims to optimize liquid vape supplies by analyzing sales patterns. This research method is K-Means Clustering which includes several stages, namely determining the number of clusters, determining the centroid point randomly, calculating the closest distance between data and the centroid using the Euclidean method, grouping data into each cluster, updating the centroid until it is stable, and evaluating the results. The data used in the research is liquid vape sales data from June to November 2024 with a total of 68 product samples. Data processing was carried out manually and testing used RapidMiner software to measure the level of accuracy of the clustering results. The research results show that the K-Means Clustering algorithm is successful in grouping products into three categories: very popular, best selling, and not very popular. 51 products are in the low-selling category, 13 products are in the best-selling category, and 4 products are in the very best-selling category, with a Davies Bouldin value of 0.374%. The application of K-Means Clustering is effective in grouping products according to demand, helps determine the ideal stock amount, reduces the risk of product excesses or shortages, and increases operational efficiency