Analisa Perbandingan Nilai Akurasi Moving Average dan Double Exponential Smoothing Brown dalam Memprediksi Kerusakan Galon Pada Distribusi AQUA PT. Tirta Investama

Abstract

A good forecasting method is a method that produces forecasts close to actual values. The Exponential Smoothing method is a frequently used forecasting method for time-series data. The Double Exponential Smoothing Brown is an exponential smoothing development method to overcome the difference between the actual data and the forecast value, if there is a trend element in the data plot. This concept is similar to the Double Moving Average method. These two methods are often used in the industrial world, to produce products promptly. As a manufacturing industry that produces water products, AQUA is not immune from various problems such as the lack of bottle availability caused by bottle damage at the distribution stage. This problem requires the company to keep investing on bottles. So that the increase in bottle damage can be more logical with more measurable causes, this study was conducted to predict gallon damage in the distribution of AQUA by PT. Tirta Investama from 3rd January to 29th December 2022, with the Moving Average and Double Exponential Smoothing Brown method. It showed that the forecast for the 13th period were obtained using the Moving Average method is 1920 gallons. Meanwhile, the results by using the Double Exponential Smoothing Brown method (α = 0.7) is 2096 gallons. The smallest MAPE calculation result is 18.44 on Moving Average method. Compared to the actual data, it can be concluded that the use of Moving Average method is more accurate than Double Exponential Smoothing Brown method in this case.