BibTex Citation Data :
@article{J.Gauss58909, author = {Hendryati Rahmi and Nur Salam and Sigit Dwi Prabowo}, title = {ANALISIS LONG MEMORY NILAI TUKAR RUPIAH TERHADAP DOLAR AMERIKA SERIKAT (USD/IDR) MENGGUNAKAN HYBRID ARFIMA-FFNN}, journal = {Jurnal Gaussian}, volume = {15}, number = {2}, year = {2026}, keywords = {ARFIMA-FFNN; Hybrid; Long Memory; USD/IDR Exchange Rate}, abstract = {The USD/IDR exchange rate is a dynamic economic indicator influenced by internal and external factors. The Autocorrelation Function (ACF) plot of daily JISDOR Bank Indonesia data for January 2023 to September 2025 indicated long memory characteristics, reflecting strong dependence between current and distant past observations, which makes the standard ARIMA model inadequate. This study aimed to develop the best forecasting model using a Hybrid ARFIMA–FFNN approach with data divided into training (80%) and testing (20%). The analysis included stationarity testing, long memory identification using Rescaled Range Statistics (R/S) and the Hurst Exponent, ARFIMA modeling, residual modeling using a Feed Forward Neural Network (FFNN) with backpropagation and Adam optimization, and accuracy evaluation using MAE and MAPE. Long memory was confirmed with Hurst Exponent H = 0,8365 and fractional parameter d = 0,3365. The best model, ARFIMA (2, 0,3365, 1)–FFNN (2-3-1), achieved a MAPE of 0,5024%, lower than standalone ARFIMA (2.6509%). The 10-business-day forecast for October 1–14, 2025 ranged from IDR 17.139,32 to IDR 17.358,10 per US dollar, indicating persistent rupiah depreciation of approximately 3–5% compared to September 2025 (IDR 16.430–IDR 16.775), with potential impacts on import costs and domestic inflation.}, issn = {2339-2541}, pages = {414--425} doi = {10.14710/j.gauss.15.2.414-425}, url = {https://ejournal3.undip.ac.id/index.php/gaussian/article/view/58909} }
Refworks Citation Data :
Article Metrics:
Last update:
The Authors submitting a manuscript do so on the understanding that if accepted for publication, copyright of the article shall be assigned to Media Statistika journal and Department of Statistics, Universitas Diponegoro as the publisher of the journal. Copyright encompasses the rights to reproduce and deliver the article in all form and media, including reprints, photographs, microfilms, and any other similar reproductions, as well as translations.
Jurnal Gaussian and Department of Statistics, Universitas Diponegoro and the Editors make every effort to ensure that no wrong or misleading data, opinions or statements be published in the journal. In any way, the contents of the articles and advertisements published in Jurnal Gaussian journal are the sole and exclusive responsibility of their respective authors and advertisers.
The Copyright Transfer Form can be downloaded here: [Copyright Transfer Form Jurnal Gaussian]. The copyright form should be signed originally and send to the Editorial Office in the form of original mail, scanned document or fax :
Dr. Rukun Santoso (Editor-in-Chief) Editorial Office of Jurnal GaussianDepartment of Statistics, Universitas DiponegoroJl. Prof. Soedarto, Kampus Undip Tembalang, Semarang, Central Java, Indonesia 50275Telp./Fax: +62-24-7474754Email: jurnalgaussian@gmail.com
Jurnal Gaussian by Departemen Statistika Undip is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Visitor Number:
View statistics