PENERAPAN DATA MINING UNTUK PREDIKSI PENJUALAN PRODUK TERLARIS MENGGUNAKAN METODE K-NEAREST NEIGHBOR (KNN) (STUDI KASUS: TOKO GROSIR RIZKI)

Riani Budiawati (Universitas Galuh)
Tuti Rohayati (Universitas Galuh)
Rian Dwicahya Supriatman (Universitas Galuh)

Jurnal Mahasiswa Sistem Informasi Galuh (JMSIG)
Vol 2 No 2 (2026): Journal of Galuh Information Systems Student

This study aims to address inventory management problems at Rizki Wholesale Store, which still relies on manual record-keeping. The lack of an accurate prediction system often leads to overstocking of certain items or stock shortages for products that are in high demand by consumers. By applying data mining techniques using the K-Nearest Neighbor (KNN) method, this study processes historical transaction data from January to December 2024 to classify 50 types of products into the categories of “Best-Selling” and “Non-Best-Selling.” The analysis results indicate that the KNN algorithm is capable of providing accurate classification based on the proximity distance between sales data points. These findings offer strategic guidance for the store owner in managing inventory procurement, particularly for essential products such as food items and cleaning supplies, which consistently demonstrate high sales performance.