Prediksi Nilai Tukar USD Terhadap Rupiah Menggunakan Metode Double Exponential Smoothing Dan Long Short-Term Memory

Sari, Dewi (2026) Prediksi Nilai Tukar USD Terhadap Rupiah Menggunakan Metode Double Exponential Smoothing Dan Long Short-Term Memory. Bachelor thesis, Institut Teknologi Kalimantan.

	Text 17221029_cover.pdf Restricted to Repository staff only until 4 October 2028. Download (598kB) | Request a copy
	Text 17221029_statement_of_authenticity.pdf Restricted to Repository staff only until 4 October 2028. Download (181kB) | Request a copy
	Text 17221029_publishing_agreement.pdf Restricted to Repository staff only until 4 October 2028. Download (210kB) | Request a copy
	Text 17221029_approval_sheet.pdf Restricted to Repository staff only until 4 October 2028. Download (271kB) | Request a copy
	Text 17221029_preface.pdf Restricted to Repository staff only until 4 October 2028. Download (257kB) | Request a copy
	Text 17221029_abstract_id.pdf Restricted to Repository staff only until 4 October 2028. Download (273kB) | Request a copy
	Text 17221029_abstract_en.pdf Restricted to Repository staff only until 4 October 2028. Download (227kB) | Request a copy
	Text 17221029_table_of_content.pdf Restricted to Repository staff only until 4 October 2028. Download (336kB) | Request a copy
	Text 17221029_illustrations.pdf Restricted to Repository staff only until 4 October 2028. Download (227kB) | Request a copy
	Text 17221029_tables.pdf Restricted to Repository staff only until 4 October 2028. Download (254kB) | Request a copy
	Text 17221029_notations.pdf Restricted to Repository staff only until 4 October 2028. Download (326kB) | Request a copy
	Text 17221029_chapter_1.pdf Restricted to Repository staff only until 4 October 2028. Download (368kB) | Request a copy
	Text 17221029_chapter_2.pdf Restricted to Repository staff only until 4 October 2028. Download (676kB) | Request a copy
	Text 17221029_chapter_3.pdf Restricted to Repository staff only until 4 October 2028. Download (381kB) | Request a copy
	Text 17221029_chapter_4.pdf Restricted to Repository staff only until 4 October 2028. Download (961kB) | Request a copy
	Text 17221029_conclusions.pdf Restricted to Repository staff only until 4 October 2028. Download (279kB) | Request a copy
	Text 17221029_bibliography.pdf Restricted to Repository staff only until 4 October 2028. Download (255kB) | Request a copy
	Text 17221029_enclosure.pdf Restricted to Repository staff only until 4 October 2028. Download (1MB) | Request a copy
	Text 17221029_paper.pdf Restricted to Repository staff only until 4 October 2028. Download (714kB) | Request a copy
	Text 17221029_presentation.pdf Restricted to Repository staff only until 4 October 2028. Download (1MB) | Request a copy
	Text 17221029_Form. TA-020.pdf.pdf Restricted to Repository staff only until 4 October 2028. Download (275kB) | Request a copy

Abstract

The USD/IDR exchange rate is a key indicator in the Indonesian economy because it affects inflation, international trade, investment, and economic stability. Exchange rate fluctuations, which are influenced by various factors, lead to high uncertainty, making it necessary to develop forecasting methods capable of producing accurate estimates. This study aims to compare the performance of the DES and LSTM methods in predicting the USD/IDR exchange rate. The data used consists of daily USD/IDR exchange rate data for the period January 2021-December 2025. The research stages included dividing the data into 80% training data and 20% test data, modeling the DES using the Holt approach with parameter optimization and via grid search, and modeling the LSTM through baseline and hyperparameter tuning stages. The performance of both models was evaluated using MAPE. The results show that the best DES model was obtained with parameters = 0.1 and = 0.3, yielding a MAPE of 0.97%, while the best LSTM model used a sliding window of 63, 32 neurons, a learning rate of 0.01, 100 epochs, a batch size of 32, a dropout rate of 0.2, and the Adam optimizer, yielding a MAPE of 0.20%. Although both methods exhibit excellent accuracy, the LSTM model produces lower prediction errors compared to DES. Exchange rate predictions for 2026 were subsequently made using the LSTM model and presented in the form of a 95% Bootstrap Prediction Interval.

Item Type:	Thesis (Bachelor)
Subjects:	A General Works > AI Indexes (General)
Divisions:	Jurusan Matematika dan Teknologi Informasi > Ilmu Aktuaria
Depositing User:	Dewi Leonita Sari
Date Deposited:	13 Jul 2026 06:27
Last Modified:	13 Jul 2026 06:27
URI:	http://repository.itk.ac.id/id/eprint/26088

Actions (login required)

View Item