EUVIMEDEUVIMED entdecken
Crossref

Feature Extraction Implementation in the Forecasting Method to Predict Indonesian Oil and Gas Exports and Imports

Michael Anggun Kado Pradana, Irfan Pratama

Edu Komputika Journal · 2024

DOI: 10.15294/edukom.v11i1.7879 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Edu Komputika Journal, 2024. Verfasst von Michael Anggun Kado Pradana und 1 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Future export and import predictions can use data mining and forecasting applications of data mining. Then, normalisation is carried out using datasets taken at the centre of the statistical agency using a mix-max scaler. The normalisation results are then calculated using several forecasting methods, such as Exponential Smoothing, SARIMAX, XGBoost, and CatBoost. The accuracy of this method can be improved by using feature extraction decomposition. They are decomposing, such as trend, residue, and seasonal. The results of the decomposition then become new features that are entered into the prediction model. The prediction results are evaluated using the root mean square error (RMSE). The smaller the RMSE, the better the results. The prediction results without using the method obtained by the Exponential Smoothing method have the best level of accuracy with an average RMSE value of 0.111 and the SARIMAX method with an average RMSE value of 0.146. Meanwhile, the prediction results using the CatBoost and XGBoost feature extraction methods have the best level of accuracy with an RMSE value of 0.046. From the results of the comparison of predictions, the addition of decomposition features to most forecasting methods can significantly increase the accuracy of the calculation.

Zitieren

APA

Pradana, M. A. K., & Pratama, I. (2024). Feature Extraction Implementation in the Forecasting Method to Predict Indonesian Oil and Gas Exports and Imports. Edu Komputika Journal. https://doi.org/10.15294/edukom.v11i1.7879

Vancouver

Pradana MAK, Pratama I. Feature Extraction Implementation in the Forecasting Method to Predict Indonesian Oil and Gas Exports and Imports. Edu Komputika Journal. 2024. doi:10.15294/edukom.v11i1.7879

BibTeX
@article{Pradana2024,
  author = {Pradana, Michael Anggun Kado and Pratama, Irfan},
  title = {Feature Extraction Implementation in the Forecasting Method to Predict Indonesian Oil and Gas Exports and Imports},
  journal = {Edu Komputika Journal},
  year = {2024},
  doi = {10.15294/edukom.v11i1.7879}
}
Kostenlos starten →