Grouping Building Material Sales for Management Optimization with the K-Means Algorithm

Abstract

Sales transactions at the Pedi Building Store are recorded in the daily logbook provided by the store, so that it cannot be immediately known which types of products are selling well or which are not selling well. To overcome this problem, an analysis of sales data is carried out using the K-Means Algorithm. The K-Means algorithm is an algorithm that requires as many input parameters and divides a set of objects into clusters so that the level of similarity between members in one cluster is high, while the level of similarity with members in other clusters is very low. This study aims to determine the types of products that sell well and don't sell well. The data used in this analysis are 100 data samples with 2 clusters formed, namely unsold goods (Cluster 1) and salable goods (Cluster 2). With the analysis of the K-Means Algorithm, it produces 4 data items that are not selling well (Cluster 1), while 96 data items that sell well (Cluster 2). With the results of the analysis of the K-Means Algorithm, a sales application was developed for the Pedi Building Store which is managed by the analysis of the K-Means Algorithm. The application makes it easy for sellers to manage data to form sales processes of salable and less salable goods