Fitriana, Ika Nur Laily (2022) Peramalan Harga Batubara Acuan (Hba) Menggunakan Model Hybrid Fungsi Transfer-Machine Learning Dengan Variabel Eksogen. Masters thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 6003201001-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6003201001-Master_Thesis.pdf Restricted to Repository staff only Download (4MB) | Request a copy |
Abstract
Studi tentang peramalan harga batubara sangat penting untuk dilakukan, karena untuk membantu perencanaan di masa yang akan datang. Harga yang menjadi acuan bagi produsen batubara di Indonesia adalah harga batubara acuan (HBA) yang dikeluarkan oleh Kementerian ESDM. Pola data HBA dari waktu ke waktu cenderung rumit, karena batubara sebagai komoditas energi utama sangat riskan dipengaruhi oleh beberapa faktor. Pada penelitian ini dikembangkan metode hybrid time series model agar peramalan periode mendatang lebih akurat. Pada peramalan HBA digunakan lima variabel eksogen. Selain menggunakan variabel lain sebagai variabel eksogen, pada penelitian ini juga akan dilakukan pemodelan tanpa variabel eksogen sebagai pembanding. Penggunaan metode hybrid dan variabel eksogen diharapkan dapat menangkap fenomena yang terjadi pada pola data sehingga menghasilkan peramalan yang akurat. Hasil dari penelitian ini yaitu model hybrid Fungsi Transfer-NN merupakan metode terbaik dibandingkan model lain yang melibatkan variabel eksogen dalam peramalan HBA. Variabel eksogen yang paling sesuai untuk meramalkan HBA secara simultan menggunakan Fungsi Transfer Multi Input yaitu harga gas alam dunia (X2,t) dan Monthly Average Rates USD ke IDR (X5,t). Perbandingan model menghasilkan kesimpulan bahwa model terbaik dalam meramalkan Harga Batubara Acuan (HBA) Indonesia adalah model tanpa melibatkan variabel eksogen, yakni model hybrid ARIMA-NN.
====================================================================================================================================
The study of coal price forecasting is very important to do, because it helps planning in the future. The reference price for coal producers in Indonesia is the reference coal price (HBA) issued by the Ministry of Energy and Mineral Resources. The pattern of HBA data from time to time tends to be complicated, because coal as the main energy commodity is very susceptible to being influenced by several factors. In this study, a hybrid time series model was developed to make forecasting of future periods more accurate. In the HBA forecasting, five exogenous variables were used. In addition to using other variables as exogenous variables, this study will also do modeling without exogenous variables as comparisons. The use of hybrid methods and exogenous variables is expected to capture phenomena that occur in data patterns so as to produce accurate forecasts. The result of this research is that the Transfer Function-NN hybrid model is the best method compared to other models that involve exogenous variables in HBA forecasting. The most suitable exogenous variables to predict HBA simultaneously using the Multi Input Transfer Function are world natural gas prices (X2,t) and Monthly Average Rates USD to IDR (X5,t). Comparison of models with exogenous variables and without exogenous variables results in the conclusion that the best model in predicting Indonesia's Reference Coal Price (HBA) is a model without involving exogenous variables, namely the ARIMA-NN hybrid model.
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTSt 519.535 Fit p-1 2022 |
| Uncontrolled Keywords: | Fungsi Transfer, Harga Batubara, Variabel Eksogen, Model Hybrid Time series, Transfer Function, Coal Price, Exogenous Variable, Hybrid Time series Model |
| Subjects: | Q Science > QA Mathematics > QA280 Box-Jenkins forecasting |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis |
| Depositing User: | Mr. Marsudiyana - |
| Date Deposited: | 27 Apr 2026 08:57 |
| Last Modified: | 27 Apr 2026 08:57 |
| URI: | http://repository.its.ac.id/id/eprint/132924 |
Actions (login required)
 |
View Item |