Grey Double Exponential Smoothing Dengan Optimasi Levenberg-Marquardt Untuk Peramalan Volume Penumpang Di Bandara Soekarno-Hatta

Abstract

Aircraft has became the best choice for long distance traveling because it has shortest travel time than any other transportations. Moreover, in recent years, aviation industries have competed for providing low cost flight so that it can also be enjoyed by middle class society. Thus escalate the popularity of aircraft as economical carrier. Knowing the volume of passengers in advance will help government and related institutions to effectively providing facilities. The volume of passengers can be predicted using classic model such as double exponential smoothing model which is simpler and has high accuracy. However, the randomness of Indonesian passenger volume data cause double exponential smoothing (DES) cannot follow both data pattern and data trend. Moreover, classic model often encounters overfitting where the prediction is bigger than the actual data. Therefore, we employed Grey Method applied on DES (GDES) to overcome this problem. GDES enabled the researcher to perform better data fitting because it would generate smoothing curve which showed clearer trend. As the result, although GDES fitting curve had higher error measurement (MSE) than DES, the forecasting result of GDES was more precise than DES. Keyword: Double Exponential Smoothing, Grey Method, Levenberg-Marquardt