Peramalan Volume Muat Kargo Domestik pada Empat Bandara Utama di Indonesia dengan Menggunakan Metode VARIMA, MLS-SVR, dan Hybrid VARIMA-MLS-SVR

Yunita, Nikita Amanda (2024) Peramalan Volume Muat Kargo Domestik pada Empat Bandara Utama di Indonesia dengan Menggunakan Metode VARIMA, MLS-SVR, dan Hybrid VARIMA-MLS-SVR. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003201150-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Transportasi adalah salah satu aspek penting dalam kehidupan manusia. Perkembangan transportasi sangat pesat dari jaman ke jaman. Salah satu transportasi yang berkembang adalah transportasi udara. Dengan adanya online shop kargo udara ini sangat diperlukan untuk mengantarkan kargo sehingga dari waktu ke waktu volume muat kargo pun berubah ubah. Perubahan volume muat kargo inilah yang membuat peneliti ingin meramalkan volume muat kargo domestik di empat bandara utama di Indonesia. Volume muat kargo antara bandara satu dengan yang lainnya saling berhubungan. Hal ini dikarenakan adanya transit kargo pada bandara satu dengan yang lain. Sehingga diperlukan metode peramalan multivariate time series dengan menentukan model terbaik. Pada penelitian ini digunakan model multivariat konvensional yaitu Vector Autoregressive Integreted Moving Avarage (VARIMA) dan pendekatan machine learning dengan metode Multioutput Least Square Support Vector Regression (MLS-SVR). Hasil penelitian menunjukkan bahwa residual model VARIMA hanya white noise hingga lag-11, sehingga dikembangkan model Hybrid VARIMA-MLS-SVR. Namun dari ketiga model tersebut, model yang terbaik untuk meramalkan volume muat kargo domestik di empat bandara di Indonesia adalah VARIMA(3,1,0) dengan nilai MAPE out sample sebesar 10,12% dan Hybrid VARIMA-MLS-SVR dengan nilai MAPE out sample sebesar 10,24%. Selain itu, model Hybrid VARIMA-MLS-SVR unggul bedasarkan kriteria MAPE in sample dengan hyper-parameter terbaik adalah γ=2^20, γ"=2^20,σ=2^20.
================================================================================================
Transportation is one of the crucial aspects of human life. The development of transportation has been very rapid over time. One of the modes of transportation that has evolved is air transportation. With the rise of online shopping, air cargo has become essential for delivering goods, leading to fluctuations in cargo volumes over time. These fluctuations in cargo volumes are what prompted researchers to forecast domestic cargo volumes at four major airports in Indonesia. The cargo volumes between these airports are interconnected due to cargo transits from one airport to another. Therefore, a multivariate time series forecasting method is needed to determine the best model. This study used a conventional multivariate model, namely the Vector Autoregressive Integrated Moving Average (VARIMA), and a machine learning approach with the Multioutput Least Square Support Vector Regression (MLS-SVR) method. The results showed that the residuals of the VARIMA model were only white noise up to lag-11, leading to the development of a hybrid VARIMA-MLS-SVR model. However, among the three models, the best model for forecasting domestic cargo volumes at the four major airports in Indonesia is VARIMA(3,1,0) with an out-of-sample MAPE value of 10.12% and Hybrid VARIMA-MLS-SVR with an out of sample MAPE value of 10.24%. Additionally, the MLS-SVR model excels based on the in-sample MAPE criterion with the best hyper-parameters being γ=2^20, γ"=2^20,σ=2^20.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	VARIMA, MLS-SVR, Hybrid VARIMA-MLS-SVR, Volume Muat Kargo, Bandara, VARIMA, MLS-SVR, Hybrid VARIMA-MLS-SVR, Volume of cargo , Airport.
Subjects:	H Social Sciences > HA Statistics > HA30.3 Time-series analysis H Social Sciences > HB Economic Theory > Economic forecasting--Mathematical models.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Nikita Amanda Yunita
Date Deposited:	12 Aug 2024 05:26
Last Modified:	12 Aug 2024 05:26
URI:	http://repository.its.ac.id/id/eprint/114950

Actions (login required)

View Item