Implementation of the K-Means Algorithm in Sales Clustering at a Company using the KDD Methodology

Abstract

This research aims to implement K-Means algorithm in sales clustering at PT Sila Tirta Gemilang using Knowledge Discovery in Databases (KDD) methodology. PT Sila Tirta Gemilang is a company operating in the bottled drinking water industry sector. This research was conducted using a KDD approach that involves collecting historical sales data and has the main objective of improving the company's understanding of their product sales patterns. K-Means Clustering algorithm is used to classify products based on similar sales characteristics. In the K-Means method, the optimal cluster center point is determined to group products with comparable sales performance. By applying clustering using K-Means algorithm and KDD method, clustering of water types that are in significant demand at PT Sila Tirta Gemilang was conducted. As a result, three clusters were found, each containing water types with different characteristics. Cluster 0 has 1 water type with a high level of interest, while Cluster 1 has 3 water types with a low level of interest. Finally, Cluster 2 consists of 2 water types with a medium level of interest. From the results that have been obtained, companies can take more appropriate steps to increase profits and optimize their sales performance.