Pengembangan Model Time Series untuk Peramalan Harga Beras Medium Eceran di Provinsi Gorontalo

Amalan Fadil Gaib (Universitas Negeri Gorontalo)
Abdul Rasyid (Universitas Negeri Gorontalo)
Jamal Darusalam Giu (Universitas Negeri Gorontalo)

Jurnal Teknik Industri Terintegrasi (JUTIN)
Vol. 9 No. 2 (2026): April

This study aims to forecast the retail price of medium-grade rice in Gorontalo Province using three Time Series models: ARIMAX, SARIMAX, and Prophet, with an exogenous variable representing the main harvest season. Monthly data from 2014–2024 were obtained from the Central Statistics Agency (BPS) of Gorontalo Province and processed using Python. Model accuracy was evaluated using Root Mean Squared Error (RMSE) and Mean Absolute Percentage Error (MAPE) on the test dataset. Results show that Prophet achieved the best performance (RMSE IDR 381.48; MAPE 0.02%), followed by ARIMAX and SARIMAX. Prophet effectively captured long-term trends and seasonal patterns, projecting price fluctuations with significant increases in mid and late 2025. These findings are expected to serve as a reference for local governments and market stakeholders in supply planning and rice price stabilization strategies.