Prediksi Penjualan Pada PT. Sewu Segar Nusantara Dengan Algoritma K-Nearest Neighbor (KNN)

Abstract

In the business world, sales analysis and prediction play a crucial role in optimizing company operations. As a fruit distributor, which is one of the business entities facing unique challenges in inventory management and meeting customer demands. Making informed decisions in managing timely fruit stock can help the company avoid excessive storage costs and inventory shortages that could adversely affect the business. K-Nearest Neighbor (KNN) is one of the widely used machine learning algorithms across various fields, including business analysis. This algorithm focuses on grouping data based on feature similarity. In the context of sales prediction, KNN can be employed to forecast fruit demand by considering patterns of previous customer purchases and other features. From the test results, it can be observed that there is an average accuracy value of 86%, recall of 87%, precision of 81%, and an f1-score of 84%. These accuracy results indicate that the created model is highly effective in predicting fruit sales at PT. Sewu Segar Nusantara.