Prediksi Penjualan Minuman Kopi Terlaris di Cafe Hidden Place dengan Pendekatan Algoritma K-Nearest Neighbor dan Naïve Bayes

Abstract

In the increasingly fierce competition for the coffee shop business, the use of data is the key to success. Cafe Hidden Place, one of the popular coffee shops in Medan city, faced the challenge of being able to adapt in order to survive and grow also increasing marketing. This research aims to improve the operational efficiency and competitiveness of Cafe Hidden Place through the use of historical sales data. Using the K-Nearest Neighbor (K-NN) and Naïve Bayes algorithms, the study successfully predicted the best-selling coffee drinks based on historical data. The prediction results show that hidden taste, tira miss u, and crème brulee drinks are the best-selling coffee drinks. The accuracy of the prediction using the RapidMiner software was obtained with an accuracy value of 100% for the K-NN algorithm while the accuracy value for the Naïve Bayes algorithm was 76.47%. With these prediction results, Cafe Hidden Place can create more targeted marketing strategies. This research proves that the use of data in the culinary business can improve decision-making and overall business performance.