IMPLEMENTASI METODE TRIPLE EXPONENTIAL SMOOTHING (BROWN) UNTUK PREDIKSI PENJUALAN BARANG LIQUID FREEBASE DAN SALT DI CV. GRESSVAPE BALONGPANGGANG

Abstract

This research was conducted to find out about sales predictions with the reason that this research was carried out as a support for the progress of CV.Gressvape income in Balongapanggang so that it could be managed to predict the amount of profits and losses in the future by applying the Triple Exponential Smoothing (Brown) method. It is a linear exponential smoothing that can be used when predicting data using the basic trend pattern, the high smoothing type can be used if the basic pattern is cubic, quadratic, or higher order. By applying this method, the number of stock items will be more transparent for the next period. In the trial of 2 items for the addition of liquid freebase and salt items in the 2nd week of August 2019 - 4th week of May 2020 there was a trial using 9 different alpha values, namely from alpha 0.1 - 0.9 and refers to a period of 1 week, 2 weeks, 3 weeks, 4 weeks earlier. In the total prediction to be a comparative value in the real data, the sum of sales of liquid freebase and salt goods, so that the prediction error value is determined using Mean Absolute Deviation (MAD) and Mean Absolute Percentage Error (MAPE). On the average of the four prediction trials, the same value was obtained from the lowest MAPE with reference to 3 weeks at a higher alpha value of 0.2.