<![CDATA[Data mining can be used to find solutions in making sales decisions to increase sales. Sales data storage stores many sales transaction records, where each document provides products purchased by customers in each sales transaction. A problem began to arise with an excess stockpiling of materials. The number of fluctuating sales causes the stock of available materials to be unstable and can directly impact consumers. Mistakes in predicting sales caused the coffee shop to buy large quantities of material stock, which were not widely used or sold out, so the supply of these materials swelled in the warehouse. One way to be implemented is by applying data mining because there are ways and methods to meet needs, one of which is the need for extensive information, then the information that we can use to determine quality in determining a decision. Therefore, it is hoped that this research can help Dpom Coffee minimize material stock inventory management cases such as shortages and excesses and make policies to increase sales by grouping menus based on sales levels using the K-means algorithm. Based on the results of processing the sales dataset at Dpom Coffee, it produces 3 clusters, namely Cluster 1 with eight menus with low sales levels, cluster 2 with 40 menus with moderate sales levels, and cluster 3 with seven menus with high sales levels. The accuracy or performance of the k-means algorithm results in a Davies Bouldin index value of 0.457.]]>
Copyrights © 2022
