PREDIKSI JUMLAH PENERBANGAN AIRFAST INDONESIA MENGGUNAKAN METODE DOUBLE EXPONENTIAL SMOOTHING

Abstract

Forecasting the number of flights is essential to support operational planning in the charter aviation industry, where demand tends to fluctuate and follow a trend pattern. Uncertainty in flight volume may affect scheduling efficiency and resource allocation. This study aims to forecast the monthly number of flights using the Double Exponential Smoothing method. The data used are monthly flight data from January 2022 to June 2025. Double Exponential Smoothing is applied to capture the level and trend components of the time series. The smoothing parameters are determined using a grid search approach with Python software to obtain the optimal model. Model accuracy is evaluated using Mean Absolute Scaled Error (MASE). The results show that the optimal Double Exponential Smoothing model is obtained with smoothing parameters α = 0.1 and β = 0.1, producing a MASE value of 0.4879. The results show that the obtained MASE value is less than one, indicating that the proposed model outperforms the naive forecasting model. The forecasting results indicate a gradual increasing trend in the number of flights, suggesting that Double Exponential Smoothing provides reliable forecasts for short- to medium-term operational planning.

Keywords: Double Exponential Smoothing, Mean Absolute Scaled Error, number of flights, time series forecasting.