Prediksi Inflasi Indonesia Berdasarkan Kurs USD/IDR Menggunakan Bagging Stacked LSTM

Kristanaya, Mirechelin (2026) Prediksi Inflasi Indonesia Berdasarkan Kurs USD/IDR Menggunakan Bagging Stacked LSTM. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) Cover Skripsi.pdf Download (1MB)
	Text (BAB I) BAB I.pdf Download (259kB)
	Text (BAB II) BAB II.pdf Restricted to Repository staff only until 13 July 2029. Download (904kB)
	Text (BAB III) BAB III.pdf Restricted to Repository staff only until 13 July 2029. Download (733kB)
	Text (BAB IV) BAB IV.pdf Restricted to Repository staff only until 13 July 2029. Download (2MB)
	Text (BAB V) BAB V.pdf Download (227kB)
	Text (Daftar Pustaka) DAFTAR PUSTAKA.pdf Download (197kB)
	Text (Lampiran) LAMPIRAN.pdf Restricted to Repository staff only until 13 July 2029. Download (289kB)

Abstract

Inflation is one of the key macroeconomic indicators that plays an important role in maintaining Indonesia's economic stability. Inflation movements are influenced by various factors, including the exchange rate of the Indonesian Rupiah against the United States Dollar (USD), the Bank Indonesia interest rate (BI Rate), and the money supply (M2). The relationships among these economic variables tend to be complex and nonlinear, making it necessary to employ forecasting methods capable of capturing data patterns more accurately. This study aims to forecast Indonesia's inflation using the Bagging Stacked Long Short-Term Memory (Bagging Stacked LSTM) method, which combines the Bootstrap Aggregating (Bagging) technique with the Stacked LSTM architecture. The data used in this study consist of monthly time series data from 2010 to 2025, with inflation as the dependent variable and the USD/IDR exchange rate, BI Rate, and M2 as independent variables. The research stages include data preprocessing, Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF) analysis, data normalization, sequence generation, hyperparameter tuning, model training, and forecasting. The tuning results indicate that the best parameters are a time step of 8, a learning rate of 0.001, 8 units in the first layer, 4 units in the second layer, a dropout rate of 0.3, a batch size of 16, and 75 epochs. The Bagging Stacked LSTM model was constructed using 5 estimators with a block size of 12 in the Moving Block Bootstrap process. The evaluation results show that the model achieved an MSE of 0.1727, an MAE of 0.3187, an RMSE of 0.4156, and a MAPE of 2.75%, indicating a low prediction error rate. The forecasting results suggest that Indonesia's inflation is expected to increase gradually in future periods, reaching a highest predicted value of 1.083652 in the 12th period. Therefore, the Bagging Stacked LSTM method is capable of providing accurate inflation forecasts and has the potential to serve as a decision-support tool for policymakers.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Damaliana, Aviolla Terza	NIDN0002089402	aviolla.terza.sada@upnjatim.ac.id
Thesis advisor	Wara, Shindi Shella May	NUPTK1850774675230252	shindi.shella.fasilkom@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA76.6 Computer Programming

Divisions:

Faculty of Computer Science > Departemen of Data Science

Depositing User:

Mirechelin Kristanaya

Date Deposited:

13 Jul 2026 06:46

Last Modified:

13 Jul 2026 06:46

URI:

https://repository.upnjatim.ac.id/id/eprint/55095