Comparison of the accuracy of seasonal data prediction values using SARIMA and winter exponential smoothing on the number of ship passengers in Batam City

Abstract

Sea transportation is one of the keys to the success of the tourism industry in a region. One of the problems faced in sea transportation is the increase in the number of passengers that is not balanced with the number of ships available. An accurate prediction model is very important in forecasting the number of passengers of a ship in order to prepare the number of ships according to the number of passengers. This study aims to compare the accuracy of prediction models on the number of ship passengers in Batam City whose data is fluctuating and contains seasonal patterns. In this study, secondary data was used in the form of the number of ship passengers in Batam City from January 2014 to December 2022. The methods used are SARIMA and Winter Exponential Smoothing. The results of this study show that the selected SARIMA model in forecasting the number of passengers in Batam City is SARIMA is SARIMA (0,1,1)(0,1,1)12 with an MSD value of 0.0000021. In contrast, the selected Winter Exponential Smoothing model is Winter Exponential Smoothing at alpha = 0.9, beta = 0.1, and delta = 0.1 with an MSD value of 0.00000. Based on these two models, the most accurate prediction model with the lowest MSD value in forecasting the number of ship passengers in Batam City is the Winter Exponential Smoothing model with values alpha = 0.9, beta = 0.1, and delta = 0.1.

Keywords

• sarima, winter exponential smoothing, forecasting