Penerapan Data Mining Memprediksi Penjualan Obat Menggunakan Metode K-Nearest Neighbor (Studi Kasus : Apotek Difana)

Abstract

The development of information technology in the digital era has provided opportunities for various sectors, including the pharmaceutical industry, to improve operational efficiency and service quality. One of the main challenges faced by pharmacies is inventory management, where both stockouts and overstocks often occur due to the limitations of conventional prediction methods. This study aims to apply data mining techniques using the K-Nearest Neighbor (KNN) algorithm to predict drug sales at Apotik Difana. KNN is chosen because it is simple yet effective in recognizing sales patterns from historical data. A web-based prediction system was developed to facilitate accessibility and usability for the pharmacy. The scope of this study focuses on historical sales data without considering external factors, and only KNN is used without comparison to other algorithms. The results are expected to assist the pharmacy in determining the right type and quantity of drugs, optimizing inventory management, reducing losses from expired drugs, and improving customer service quality. Furthermore, this research provides a theoretical contribution to the development of data mining in sales prediction and offers a practical, technology-based solution for pharmaceutical inventory management.