Implementasi Metode Pemulusan Eksponensial Ganda Brown dan Holt pada Peramalan Jumlah Penduduk di Kota Medan Tahun 2023-2025

Abstract

Forecasting is an activity carried out to estimate or predict an event in the future. Forecasting can be done using several methods. One Parameter Double Exponential smoothing method from Brown and Two Parameters from Holt is a smoothing method commonly used in forecasting with linear trend data patterns. In making a forecast, the level of error (error) of a method greatly affects the level of forecasting accuracy. The smaller the error rate, the forecast is said to be better and vice versa if the error rate is greater, then the forecast needs to be evaluated. The error rate of a forecast can be measured using the Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). In this study, forecasting the population of the city of Medan was carried out using the double exponential smoothing method with one parameter from Brown and two parameters from Holt with the aim of knowing the population of Medan City in 2023-2025. Forecasting is done based on population data from 2010-2022. Based on calculations carried out using the Brown method using the parameter α = 0.5, the forecast for the population of Medan City in 2023 will be 2,548,677 people, an increase of 2.17% from 2022 and in 2024 and 2025 each will increase by 4 .15% and 6.13% with an MAE value of 23116.790 and a MAPE of 0.9848%. Calculations using the Holt method with α=0.9 and γ=0.1 produce a forecast of the population of Medan City in 2023 of 2,524,962 people or an increase of 1.22% from 2022 and in 2024 and 2025 each increase of 2.45% and 3.69%, and the MAE value is 19587 and the MAPE is 0.8301%.