Penerapan Metode Moving Average, Exponential Smoothing, dan Linear Regression dalam Peramalan Jumlah Produksi Okky Koko Drink Leci pada PT. XYZ

Abstract

PT. XYZ is a company engaged in the soft drink manufacturing industry. One of its products is Okky Koko Drink Lychee. The production capacity of PT. XYZ is 6,000 boxes per shift. The problem that often occurs at PT. XYZ is the discrepancy between the company's work plan and budget (RKAP), production that has been determined at the beginning of the year based on actual data releases. So that the production planning that has been carried out by PPIC is considered less valid, because it is only based on historical data estimates. Based on the results of the study that has been carried out by determining data patterns using the eviews application using the Augmented Dickey Fuller (ADF) test, a trend data pattern is obtained and the methods used are Moving Average, Exponential Smoothing and Linear Regression. Using a smoothing constant with a value of M = 2-9 on the Moving Average, ? used in Exponential Smoothing 0.1-0.9 and x = 2 on Linear Regression. The method that has the smallest level of forecasting error is the Exponential Smoothing method with a value of a = 0.8, because it has a MAD value of 29,931.92, an MSE value of 1,328,504,000 and a MAPE value of 0.27% with a forecast value for the next period of 160,269. So a = 0.8 is the best for forecasting Okky Koko Drink Lychee at PT. XYZ