Komparasi Model Single Moving Avarage & Exponential Smoothing Untuk Peramalan Penjualan AMDK NuClees

Abstract

This study aims to (1) forecast the demand for NUCless bottled drinking water in the Nganjuk Regency market. (2) Find the suitable model by comparing the most accurate results from the Moving Average model and the Exponential Smoothing model. Meanwhile, to test the accuracy of the casting results, apply a control tracking signal and test the mean absolute deviation (MAD) error values, mean square error (MSE), and mean fundamental percentage error (MAPE). This research data is a time series with a seasonal pattern from January to December 2021. The results of this study place the Single Moving Average model better than the results of the Exponential Smoothing model. This is based on the results of the control tracking signal Smoothing model test results that there are deviations. In contrast, the results of the MSE, MAD, and MAPE accuracy tests of data deviations in the Single Moving Average model are lower than the results of the Exponential Smoothing model, so these results reject the results of several previous researchers who rely on the Exponential Smoothing model instead of the Exponential Smoothing model. Single Moving Average to forecast sales.   Keywords: Comparison, Moving Average Model, Exponential Smoothing Model