Klasterisasi Tingkat Penjualan pada Startup Panak.id dengan Algoritma K-Means

Abstract

Sales data processing is one way for companies to support future decision making. One way to process sales data is to cluster with data mining techniques that are present as a company solution to be able to process sales data that is owned by carrying out processes that generate new knowledge. Knowledge obtained from owned data is commonly referred to as Knowledge Discovery in Database (KDD) which consists of steps such as; data selection, data cleaning, pattern search (data mining), and interpretation. In this study, the author will discuss data mining techniques and applications using the K-Means algorithm which is applied in the Python programming language. The research was conducted at the startup company Panak.id by accessing data on sales of processed livestock products for the period July 2021 – July 2022. The results showed that the level of product sales was dominated by products that were in the less-selling category. This shows that of the many types of products offered by Panak.id, there are still many products that are of little interest to the public. From this knowledge, the company can make decisions related to sales strategies and innovations in products that are not in demand with the hope of further increasing sales. In addition, the company can also use this knowledge as a prediction in terms of stock control so that products that are not selling well can be monitored more closely to avoid stock buildup in warehouses.