Forecasting Five Indonesian Food Commodity Group Import Values using SARIMA and XGBoost

Evelin Alim Natadjaja, Trianggoro Wiradinata

Abstract


Food commodity import values exhibit diverse temporal characteristics, which may lead to variations in forecasting performance across commodity groups. This study aims to compare the performance of the Seasonal Autoregressive Integrated Moving Average (SARIMA) and eXtreme Gradient Boosting (XGBoost) algorithms in forecasting the monthly import values of five Indonesian food commodity groups classified under the Harmonized System (HS), namely HS04, HS07, HS10, HS12, and HS17. The study adopts a time-series forecasting approach following the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology, using data from Statistics Indonesia (BPS) covering the period from January 2014 to May 2026. Model configurations were selected independently for each algorithm and commodity group using expanding-window time-series cross-validation with two folds and a 12-month forecasting horizon. The selected models were subsequently evaluated using test data covering June 2025 to May 2026, based on the Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). The results indicate that neither algorithm consistently outperformed the other across all commodity groups. XGBoost achieved lower MAPE values for HS04, HS07, HS12, and HS17, whereas SARIMA performed better for HS10. The average MAPE across the five commodity groups was 23.17% for XGBoost and 31.14% for SARIMA. Both algorithms were subsequently implemented in an interactive dashboard to present model evaluation results and import value forecasts for the period from June 2026 to May 2027, enabling users to select the appropriate model according to their analytical needs.

Keywords


food import values; food commodity groups; machine learning; statistical modeling; time series forecasting

Full Text:

PDF

References


B. P. S. Indonesia, “Ekspor dan Impor Indonesia Mei 2026 masing-masing tercatat USD 23,20 miliar dan USD 24,81 miliar.” Accessed: Sep. 12, 2026. [Online]. Available: https://www.bps.go.id/id/pressrelease/2026/07/01/2587/ekspor-dan-impor-indonesia-mei-2026-masing-masing-tercatat-usd-23-20-miliar-dan-usd-24-81-miliar.html

B.-S. Indonesia, “Foreign Trade Statistical Import of Indonesia 2025.” Accessed: Sep. 12, 2026. [Online].Available: https://www.bps.go.id/en/publication/2026/08/04/29900d4157c9ba7ad7fc72ea/statistik-perdagangan-luar-negeri-indonesia-impor-2025.html

J. Tamin, B. Robiani, M. Teguh, and M. Mukhlis, “Food and Beverage Industry Sector Linkages in Indonesia,” Equity J. Ekon., Vol. 12, No. 1, pp. 83–94, Jul. 2024, DOI: 10.33019/equity.v12i1.284.

D. Nauly, Harianto, S. Hartoyo, and T. Novianti, “Kepemilikan Asing dan Kecenderungan Ekspor-Impor Industri Makanan Olahan Indonesia,” Bul. Ilm. Litbang Perdagang., Vol. 14, No. 1, pp. 47–74, Jul. 2020, DOI: 10.30908/bilp.v14i1.431.

D. D. Pane and A. A. Patunru, “The Role of Imported Inputs in Firms’ Productivity and Exports: Evidence from Indonesia,” Rev. World Econ., Vol. 159, No. 3, pp. 629–672, Aug. 2023, DOI: 10.1007/s10290-022-00476-z.

F. Petropoulos et al., “Forecasting: Theory and Practice,” Int. J. Forecast., Vol. 38, No. 3, pp. 705–871, Jul. 2022, DOI: 10.1016/j.ijforecast.2021.11.001.

J. Warnau et al., “Forecasting Industrial Production: A Comparative Study based on Volatility and Seasonality of Time Series,” Procedia Comput. SCI., vol. 277, pp. 2183–2192, 2026, DOI: 10.1016/j.procs.2026.02.256.

B. P. S. Indonesia, “Data Ekspor Impor - Badan Pusat Statistik.” Accessed: Sep. 25, 2026. [Online]. Available: https://www.bps.go.id/id/exim

M. A. Yafi, M. R. D. Maharani, N. A. Nabilla, and A. S. Adyanti, “Predictive Trends of Agricultural Food Commodities Prices in Indonesia: A Comprehensive Study using Time Series Forecasting,” Agro Ekon., Vol. 37, No. 1, p. 1, Jun. 2026, DOI: 10.22146/ae.104454.

M. Sari, S. Duran, H. Kutlu, B. Guloglu, and Z. Atik, “Various Optimized Machine Learning Techniques to Predict Agricultural Commodity Prices,” Neural Comput. Appl., Vol. 36, No. 19, pp. 11439–11459, Jul. 2024, DOI: 10.1007/s00521-024-09679-x.

M. Roznik, A. K. Mishra, and M. S. Boyd, “Using a Machine Learning Approach and Big Data to Augment WASDE Forecasts: Empirical Evidence from US Corn Yield,” J. Forecast., Vol. 42, No. 6, pp. 1370–1384, Sep. 2023, DOI: 10.1002/for.2956.

Pradana, Michael Anggun Kado and Irfan Pratama, “Feature Extraction Implementation in the Forecasting Method to Predict Indonesian Oil and Gas Exports and Imports,” Edu Komputika J., Vol. 11, No. 1, pp. 11–22, Aug. 2024, DOI: 10.15294/edukom.v11i1.7879.

Muh. A. Rasyid and N. T. Lapatta, “Development of an Indonesian Trade Forecasting Information System based on Statistical Models and Gradient Boosting,” J. Sisfokom Sist. Inf. Dan Komput., Vol. 15, No. 02, pp. 208–216, Apr. 2026, DOI: 10.32736/sisfokom.v15i02.2586.

A. I. Kurnadipare, S. Amaliya, K. A. Notodiputro, Y. Angraini, and L. N. A. Mualifah, “Comparing Forecasts of Agricultural Sector Export Values using SARIMA and Long Short-Term Memory Models,” BAREKENG J. Ilmu Mat. Dan Terap., Vol. 19, No. 1, pp. 385–396, Jan. 2025, DOI: 10.30598/barekengvol19iss1pp385-396.

G. G. Ghiffary, E. D. D. Yanuari, K. A. Notodiputro, Y. Angraini, and L. N. A. Mualifah, “Comparative Performance of SARIMAX and LSTM Model in Predicting Import Quantities of Milk, Butter, and Eggs,” BAREKENG J. Ilmu Mat. Dan Terap., Vol. 19, No. 1, pp. 407–418, Jan. 2025, DOI: 10.30598/barekengvol19iss1pp407-418.

T. Zhao, G. Chen, S. Suraphee, T. Phoophiwfa, and P. Busababodhin, “A Hybrid TCN-XGBoost Model for Agricultural Product Market Price Forecasting,” PLOS One, Vol. 20, No. 5, p. e0322496, May 2025, DOI: 10.1371/journal.pone.0322496.

M. Sahid, M. Ramdhani, and G. Darmawan, “A Hybrid Simple Moving Average and XGBoost Approach for Enhanced Wheat Commodity Price Forecasting,” Acadlore Trans. Appl. Math. Stat., Vol. 02, No. 04, pp. 238–251, Dec. 2024, DOI: 10.56578/atams020404.

J. Pardede, A. Putri Setyaningrum, M. Ilyas Al-Fadhlih, and M. Premitasari, “Impact of Feature Engineering on XGBoost Model for Forecasting Cayenne Pepper Prices,” SCI. J. Inform., Vol. 12, No. 4, pp. 651–664, Nov. 2025, DOI: 10.15294/sji.v12i4.32157.

F. H. S. Beyuta, D. G. Langamin, K. E. R. Pon, H. B. Copel, and R. G. Artes Jr., “Applications of XGBoost and SARIMA Forecasting Models on Water Consumption of MSU-TCTO: A Data-Driven Approach to Water Resource Management,” Adv. Appl. Stat., Vol. 91, No. 5, pp. 597–614, Apr. 2024, DOI: 10.17654/0972361724032.

S. S. Bora and A. L. Katchova, “Multi-Step Commodity Forecasts using Deep Learning,” Agric. Finance Rev., Vol. 84, No. 4/5, pp. 269–296, Nov. 2024, DOI: 10.1108/AFR-08-2023-0105.

J. M. Moreira, A. C. P. L. F. Carvalho, and T. Horváth, A General Introduction to Data Analytics, 1st ed. Wiley, 2018. DOI: 10.1002/9781119296294.




DOI: https://doi.org/10.32520/stmsi.v15i11.7001

Article Metrics

Abstract view : 0 times
PDF - 0 times

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.