SISTEM PREDIKSI PENJUALAN MAKANAN DAN MINUMAN PADA KAFE DENGAN MENGGUNAKAN METODE TRIPLE EXPONENTIAL SMOOTHING

Abstract

The food service operation based primarily on the sale of food and beverages. Raw material inventory provides flexibility in procurement, without sufficient inventory the company must always prepare sufficient funds for each time to buy the necessary raw materials. Many forecasting methods can be used to predict, one of which is the time series forecasting method. Time series models are used to predict the future using historical data. In other words, time series models look at what happened at a certain time and use past data to predict. This study aims to predict future sales stock using previous historical data and pay attention to the accuracy of prediction errors or MAPE at Mie Setan Noodle and Dimsum Tangerang cafe. The prediction system is built using Brown's triple exponential smoothing algorithm which has one parameter. The system for predicting stock based on sales data or historical data was successfully created. The results showed that the system built using alpha 0.5 had an error percentage of around 22%.