1.png)
Prediksi Ekspor Jasa Transportasi Indonesia Menggunakan LSTM Berbasis Data Perdagangan Global Terbuka
Sari
Penelitian ini memprediksi nilai ekspor jasa transportasi Indonesia menggunakan model Long Short-Term Memory (LSTM) berbasis data terbuka perdagangan global dalam mengatasi pola nonlinier dan ketergantungan temporal. Peneliti melatih model LSTM tiga lapis dengan aktivasi ReLU dan optimasi Adam menggunakan data ekspor tahunan (2005–2023) dari World Bank dan UNCTAD, dengan pembagian data latih-uji 80:20. Model mencapai MAPE 0,89% dan koefisien korelasi r = 0,999 (p < 0,0001), menunjukkan presisi tinggi. Model secara akurat menangkap gangguan akibat pandemi dan tren pemulihan, menawarkan alat prediksi berbasis AI untuk perencanaan ekspor dan kebijakan perdagangan. Ini merupakan studi pertama yang menerapkan LSTM pada ekspor jasa transportasi Indonesia dengan data terbuka, memberikan kontribusi metodologis dan praktis untuk negara berkembang.
Kata kunci: kecerdasan buatan, peramalan ekspor, Indonesia, LSTM, layanan transportasi
ABSTRACTThis study forecasts Indonesia’s transport service export values using a Long Short-Term Memory (LSTM) model based on open global trade data in capturing nonlinear patterns and temporal dependencies. A three-layer LSTM model is trained using ReLU activation and Adam optimization on annual export data from 2005 to 2023 sourced from the World Bank and UNCTAD. The dataset is split into 80% training and 20% testing portions. The model achieves a MAPE of 0.89% and a correlation coefficient of r = 0.999 (p < 0.0001), indicating high precision.The model accurately reflects pandemic-induced shocks and subsequent recovery trends, provides an AI-driven forecasting tool for export planning and trade policy. This is the first study to apply LSTM to Indonesia’s transport service exports using open data, contributing methodological advancement and practical value for developing economies.
Keywords: artificial intelligence, export forecasting, Indonesia, LSTM, transport services
Teks Lengkap:
PDFReferensi
Chen, Y., Wang, L., & Deng, Y. (2023). Artificial Intelligence Review, 56(2), 1119–1145. https://doi.org/10.1007/s10462-022-10122-7
Dong, Y., Wang, X., Zhang, Y., & Liu, L. (2023). Deep learning-based trade volume prediction with cross-border economic indicators. Expert Systems with Applications, 215, 119279. https://doi.org/10.1016/j.eswa.2022.119279
Guo, Z., Jiang, Y., Liu, Q., & Liang, L. (2023). Long Short-Term Memory networks for time-series forecasting in trade logistics: A hybrid framework. Journal of Intelligent & Fuzzy Systems, 44(1), 177–188. https://doi.org/10.3233/JIFS-223183
Huang, L., & Lin, M. (2020). Journal of Cleaner Production, 258, 120680. https://doi.org/10.1016/j.jclepro.2020.120680
Khan, I., Javaid, M., & Ahmed, A. (2022). Forecasting South Asian export trends using ARIMA and ARDL models: A comparative study. Asia-Pacific Journal of Regional Science, 6(2), 123–139. https://doi.org/10.1007/s41685-022-00217-z
Kou, G., Xu, Y., Peng, Y., & Shen, F. (2021). Artificial intelligence in forecasting trade flows: A review and case analysis. Technological Forecasting and Social Change, 173, 121039. https://doi.org/10.1016/j.techfore.2021.121039
Lee, D., & Kim, J. (2022). Expert Systems with Applications, 203, 117477. https://doi.org/10.1016/j.eswa.2022.117477
Li, H., & Zhao, Y. (2022). Neurocomputing, 489, 200–212. https://doi.org/10.1016/j.neucom.2021.09.078
Li, X., & Zhu, Y. (2020). Computers & Industrial Engineering, 147, 106664. https://doi.org/10.1016/j.cie.2020.106664
Luo, X., Liu, C., & Wang, Q. (2021). Knowledge-Based Systems, 233, 107550. https://doi.org/10.1016/j.knosys.2021.107550
Mehta, A., Nair, M., & Saha, A. (2022). Predicting trade dynamics using LSTM and machine learning: A data-driven approach. International Journal of Forecasting, 38(3), 912–926. https://doi.org/10.1016/j.ijforecast.2021.05.002
Naqvi, S. M., & Qureshi, F. (2021). Performance comparison of statistical and AI models for Pakistan’s exports forecasting. Economic Modelling, 94, 11–20. https://doi.org/10.1016/j.econmod.2020.10.002
Qiu, Y., Xu, J., & Wang, S. (2023). Journal of Economic Behavior & Organization, 208, 143–156. https://doi.org/10.1016/j.jebo.2023.01.008
Rahman, M. M., & Mohamad, A. (2023). Heliyon, 9(1), e12799. https://doi.org/10.1016/j.heliyon.2022.e12799
Silva, D. R., & Melo, M. T. (2021). Transportation Research Part E, 147, 102230. https://doi.org/10.1016/j.tre.2020.102230
Tang, X., & He, L. (2023). Sustainable Cities and Society, 94, 104662. https://doi.org/10.1016/j.scs.2023.104662
UNCTAD. (2023). World Trade Report. https://unctad.org
Wang, J., & Zhao, X. (2021). IEEE Access, 9, 103499–103512. https://doi.org/10.1109/ACCESS.2021.3098983
World Bank. (2023). World Development Indicators. https://data.worldbank.org
Wu, T., Chen, K., & Yan, M. (2023). IEEE Transactions on Neural Networks and Learning Systems, 34(1), 21–33.
Yang, J., Zhang, T., & Xu, L. (2022). Transport Policy, 116, 113–125. https://doi.org/10.1016/j.tranpol.2021.10.011
Zhai, Y., Chen, X., & Lin, Z. (2022). Information Sciences, 585, 87–102. https://doi.org/10.1016/j.ins.2021.11.024
Zhang, B., & Liu, J. (2020). SpringerPlus, 9, 1002–1015.
Zhang, Y., & Lin, C. (2021). Deep learning for macroeconomic forecasting: Evidence from China’s trade data. Applied Economics Letters, 28(17), 1524–1528. https://doi.org/10.1080/13504851.2020.1819774
Zhou, Y., Ma, Y., & Li, F. (2022). Decision Support Systems, 158, 113775. https://doi.org/10.1016/j.dss.2022.113775
DOI: https://doi.org/10.26760/mindjournal.v10i2.130-144
Refbacks
- Saat ini tidak ada refbacks.
____________________________________________________________
ISSN (Print): 2338-8323 | ISSN (Online): 2528-0902
Dipublikasikan oleh:
Program Studi Informatika, Institut Teknologi Nasional Bandung
Alamat:
Gedung 2 Informatika, Jl. PHH Mustofa No. 23, Bandung 40124, Indonesia
Kontak:
Telp: +62-22-7272215 (ext. 181) Fax: +62-22-7202892
Email: mind.journal@itenas.ac.id
______________________________
Statistik Pengunjung :
Jurnal ini terlisensi oleh Creative Commons Attribution-ShareAlike 4.0 International License.
