<![CDATA[When shopping, buyers often have difficulty finding daily necessities. One of the causes of this is because the product arrangement process in minimarkets is still carried out randomly and does not match consumer shopping patterns. On the contrary, buyers usually want to buy products through daily necessities packages, but these packages are usually not yet available in minimarkets. Identifying relationship patterns in minimarket transaction data can help overcome product arrangement and product packaging problems. By using the clustering method, objects are grouped into groups that have many similarities with each other. This method allows the grouping process to be carried out. Some of the methods in clustering include the K-Means and K-medoids methods. The purpose of this study is to group the data on goods in the minimarket which can be a guide for more neatly arranged product planning. Data grouping is divided into 3 categories namely slow, medium and fast. The results obtained show that the two algorithms produce different Davies-Bouldin Index values, with the K Medoids algorithm obtaining a lower value of 0.50387 while K-Means obtains a value of 0.50391 where the K-Medoids clustering results have better quality compared to K-Means. With the results of the grouping of these goods data, minimarkets can balance the stock of goods to prevent excess or shortage of inventory of these goods.]]>
Copyrights © 2024
