Peramalan Nilai Ekspor-Impor di Negara-Negara ASEAN dengan Hybrid Generalized Space Time Autoregressive With Exogeneous Variable – Transformer (GSTARX-Transformer)

Sihombing, Lukas Linthong (2024) Peramalan Nilai Ekspor-Impor di Negara-Negara ASEAN dengan Hybrid Generalized Space Time Autoregressive With Exogeneous Variable – Transformer (GSTARX-Transformer). Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5003201137-Undergraduate_Thesis.pdf] Text
5003201137-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 October 2026.

Download (22MB) | Request a copy

Abstract

Kondisi ekonomi di ASEAN menunjukkan keragaman yang signifikan. Beberapa negara anggota, seperti Indonesia, telah menunjukkan pertumbuhan ekonomi yang kuat, dengan pertumbuhan PDB sebesar 5,17 persen pada kuartal II 2023. Sementara itu, integrasi ekonomi antarnegara anggota ASEAN, seperti melalui Masyarakat Ekonomi ASEAN (MEA), telah memainkan peran penting dalam menentukan arah pertumbuhan ekonomi di wilayah tersebut. Perdagangan internasional merupakan hal terpenting dari integrasi ekonomi dalam konteks globalisasi, Indonesia sendiri telah mengalami perubahan signifikan dalam ekspor dan impor akibat fluktuasi nilai tukar, perubahan harga komoditas global, dan berbagai kebijakan perdagangan internasional. Dengan kondisi ekonomi dan target dimasa depan yang dimiliki oleh ASEAN terutama Indonesia, peramalan nilai ekspor dan impor yang dipengaruhi secara spasial, diperlukan sebagai proyeksi, penentuan strategi, dan evaluasi bagi pihak yang berkepentingan. Penelitian terdahulu menunjukkan nilai ekspor dan impor dipengaruhi oleh nilai tukar uang. Studi ini berfokus pada penerapan model space-time sebagai model yang memperhatikan aspek spasial dinamika ekspor dan impor. Penelitian ini bertujuan untuk membentuk model hybrid GSTARX dan Transformer pada data space-time. Dalam penelitian ini digunakan data time series bulanan dari Januari 2002 – Desember 2023. Hasil dari penelitian ini adalah model terbaik untuk data ekspor-impor di negara-negara ASEAN yang didapat adalah Model Hybrid Transformer (fungsi aktivasi ReLU, dropout 0,1, density 256)- GSTARX([81]) dan Transformer-GSTARX([121]) dengan nilai MAPE out-sample berturut – turut sebesar 6,46% dan 5,33%. Dengan demikian model ini digunakan untuk meramal nilai ekspor-impor di negara-negara ASEAN.
============================================================
Economic conditions in ASEAN show significant diversity. Several member countries, such as Indonesia, have shown strong economic growth, with GDP growth of 5.17 percent in the second quarter of 2023. Meanwhile, economic integration between ASEAN member countries, such as through the ASEAN Economic Community (AEC), has played an important role in determining the direction of economic growth in the region. International trade is the most important thing for economic integration in the context of globalization. Indonesia itself has experienced significant changes in exports and imports due to exchange rate fluctuations, changes in global commodity prices, and various international trade policies. With the economic conditions and future targets that ASEAN, especially Indonesia has, spatially influenced forecasting export and import values is needed as a projection, strategy determination and evaluation for interested parties. Previous research shows that the value of exports and imports is affected by the money exchange rate. In this research, the space-time model is applied as a model that takes into spatial aspects dynamics of exports out and imports from outside. This research aims to form a hybrid GSTARX and Transformer model on space-time data. In this study, monthly time series data from January 2002 - December 2023 were used. The results of this study were that the best model for export-import data in ASEAN countries was the Transformer (ReLU activation function, dropout 0,1, density 256)-GSTARX([81]) Hybrid Model and Transformer-GSTARX([121]) with out-sample MAPE values of 6,46% and 5,33% respectively. Thus, this model is used to predict the value of exports and imports in ASEAN countries.

Item Type: Thesis (Other)
Uncontrolled Keywords: Export-Import, Forecasting, GSTARX, Hybrid, Macroeconomy, Transformer, Ekspor-Impor, GSTARX, Hybrid, Makroekonomi, Peramalan, Transformer
Subjects: H Social Sciences > HA Statistics > HA30.3 Time-series analysis
H Social Sciences > HB Economic Theory
H Social Sciences > HB Economic Theory > Economic forecasting--Mathematical models.
H Social Sciences > HC Economic History and Conditions > HC441 Macroeconomics.
Divisions: Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Lukas Linthong Sihombing
Date Deposited: 09 Aug 2024 07:09
Last Modified: 09 Aug 2024 07:09
URI: http://repository.its.ac.id/id/eprint/115235

Actions (login required)

View Item View Item