Implementasi Machine Learning Untuk Prediksi Penjualan Oli Shell Pada CV. Harapan Karya Mandiri Bengkulu

Prediction is an activity of guessing or estimating something that will happen in the future by utilizing historical data through a scientific method. CV. Harapan Karya Mandiri in predicting oil sales is still in a conventional way. The weaknesses of conventional systems are in addition to human error in doing calculations and writing and can also be lost when doing recapitulation. For this reason, a machine learning technique is needed that is able to predict sales. One of the algorithms included in Machine Learning is K-Nearest Neighbor. The system implementation uses the PHP programming language with the MySql database and the method used in this research is the Waterfall method. The waterfall method is able to analyze the needs used to find out from the weaknesses of the old system, then make a design of the design and continue with the design of the new system. The conclusion from the results of this study explains that the sales prediction process with the K-Nearest Neighbor method first goes through a training process. The prediction results are also strongly influenced by the amount of data being trained and the value of "k" in this method.