| dc.contributor.advisor | Najib, Mohamad Khoirun | |
| dc.contributor.advisor | Khatizah, Elis | |
| dc.contributor.author | FIRDHASARI, AZIZAH AULIA | |
| dc.date.accessioned | 2026-06-30T02:23:55Z | |
| dc.date.available | 2026-06-30T02:23:55Z | |
| dc.date.issued | 2026 | |
| dc.identifier.uri | http://repository.ipb.ac.id/handle/123456789/173766 | |
| dc.description.abstract | Harga emas Indonesia berfluktuasi akibat pengaruh faktor makroekonomi global dan domestik sehingga diperlukan model prediksi yang akurat. Penelitian ini membandingkan kinerja metode SARIMAX dan LSTM dalam memprediksi harga emas Indonesia periode 2025-2026 dengan dua skenario, yaitu eksogen lengkap dan eksogen terpilih. Variabel eksogen terpilih ditentukan melalui analisis sensitivitas indeks Sobol. Data yang digunakan berupa data bulanan periode Januari 2010 hingga Desember 2024. Hasil menunjukkan bahwa eksogen terpilih dapat meningkatkan kinerja kedua metode dengan SARIMAX lebih unggul dibandingkan LSTM. Model SARIMAX(0,1,3)(0,0,0)12 ditetapkan sebagai model terbaik dengan MAPE 1.81% yang merupakan nilai terendah dibandingkan model lainnya. Analisis sensitivitasmenunjukkan bahwa harga emas dunia merupakan faktor paling berpengaruh, diikuti kurs USD/IDR dan USD Index. Proyeksi menunjukkan tren kenaikan harga emas yang relatif stabil tanpa pola musiman yang kuat. | |
| dc.description.abstract | Indonesian gold prices fluctuate due to the influence of global and domestic macroeconomic factors thus requiring an accurate prediction model. This study compares the performance of the SARIMAX and LSTM methods in predicting Indonesian gold prices for the 2025-2026 period through two scenarios, complete exogenous variables and selected exogenous variables. The selected exogenous variables are determined through a sensitivity analysis of the Sobol index. The data used are monthly data from January 2010 to December 2024. The results show that the selected exogenous variables can improve the performance of both methods with SARIMAX outperforming LSTM. The SARIMAX(0,1,3)(0,0,0)12 model is determined as the best model with a MAPE of 1.81% the lowest value among all models. The sensitivity analysis shows that the world gold prices is the most influential factor, followed by the USD/IDR exchange rate and the USD Index. The projections show a relatively stable upward trend in gold prices without a strong seasonal pattern. | |
| dc.description.sponsorship | | |
| dc.language.iso | id | |
| dc.publisher | IPB University | id |
| dc.subject.ddc | Mathematics | id |
| dc.subject.ddc | Forecasting | id |
| dc.title | Perbandingan Model SARIMAX dan LSTM dalam Memprediksi Harga Emas Indonesia dengan Faktor Makroekonomi Global dan Domestik | id |
| dc.title.alternative | Comparison of SARIMAX and LSTM Models in Forecasting Indonesian Gold Prices Using Global and Domestic Macroeconomic Factors | |
| dc.type | Skripsi | |
| dc.subject.keyword | gold price | id |
| dc.subject.keyword | LSTM | id |
| dc.subject.keyword | macroeconomics | id |
| dc.subject.keyword | prediction | id |
| dc.subject.keyword | SARIMAX | id |
| dc.subject.keyword | Sobol | id |
| dc.subtype | Undergraduate Theses | |