Retail business is the business of selling goods or services to consumers in units or retail. This retail business is part of the distribution channel that plays a vital role in a series of marketing activities as well as a liaison between the interests of producers and consumers. Based on sales transaction data in retail stores in 2020 obtained from www.kaggle.com, the inventory of goods is not proportional to the sales of goods. Excessive inventory and low sales levels resulted in goods accumulation in retail stores. When the sales cycle of goods is down, the stock must be prepared according to the level of sales. It takes a grouping of data to schedule an inventory of interests following the status of the purchase of goods. The data grouping used in this study uses the K-Medoids algorithm. K-Medoids is a method of partitioning clustering to group a set of (n) objects into several (k) clusters. Based on the elbow method, the optimal cluster number is 2 clusters. From the clustering process, the results obtained are cluster 1 has 320 data and cluster 2 has 765 data. The accuracy level of the cluster formed using the Davies-Bouldin Index method is 0.662748, and the Silhouette Coefficient is 0.276353.

Keywords

Retail Business, K-Medoids, Davies-Bouldin Index, Silhouette Coefficient.