Hafizh, Muhammad Abdullah (2025) Prediksi Kurs Rupiah Terhadap Dolar Amerika Berbasis Bidirectional Long Short Term Memory dan Attention Mechanism. Undergraduate thesis, UPN Veteran Jawa Timur.
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (Cover)
21081010099-cover.pdf Download (1MB) |
|
Text (Bab 1)
21081010099-bab1.pdf Download (219kB) |
|
Text (Bab 2)
21081010099-bab2.pdf Restricted to Repository staff only until 4 December 2027. Download (637kB) |
|
Text (Bab 3)
21081010099-bab3.pdf Restricted to Repository staff only until 4 December 2027. Download (655kB) |
|
Text (Bab 4)
21081010099-bab4.pdf Restricted to Repository staff only until 4 December 2027. Download (3MB) |
|
Text (Bab 5)
21081010099-bab5.pdf Download (209kB) |
|
Text (Daftar Pustaka)
21081010099-daftarpustaka.pdf Download (165kB) |
|
Text (Lampiran)
21081010099-lampiran.pdf Restricted to Repository staff only until 4 December 2027. Download (4MB) |
Abstract
The exchange rate is an indicator that reflects the economic condition of a country. Its fluctuations are influenced by various factors, particularly macroeconomic variables such as interest rates, inflation, exports, and imports. Exchange rate instability can affect multiple sectors, including international trade, banking, and tourism. Therefore, this study aims to predict the exchange rate of the IDR against the USD using the Bidirectional Long Short Term Memory (BiLSTM) algorithm, which is widely used for time series prediction, optimized with the Attention Mechanism. Attention Mechanism optimizes important information by assigning attention weights. The experimental results show that the BiLSTM-Attention model achieved the best performance compared to the BiLSTM and LSTM models. The BiLSTM-Attention model produced an RMSE of 54.45, an R² of 0.9718, and a MAPE of 0.26%, with a training time of 150 seconds. These findings demonstrate that the BiLSTM-Attention model provides better performance compared to the BiLSTM dan LSTM models, with an error rate that is 44.9% lower than BiLSTM and 47.1% lower than LSTM.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | T Technology > T Technology (General) | ||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Informatics | ||||||||||||
| Depositing User: | Hafizh Muhammad Abdullah | ||||||||||||
| Date Deposited: | 04 Dec 2025 04:06 | ||||||||||||
| Last Modified: | 04 Dec 2025 04:06 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/47787 |
Actions (login required)
 |
View Item |