Sistem Prediksi Harga Pangan DI Kabupaten Bengkalis Berbasis Web Menggunakan Metode SARIMA

Fluctuations in food prices such as chili, onion, and other staple commodities in Bengkalis Regency often occur due to the influence of weather, distribution, supply, and increased demand during certain periods such as religious holidays. Historical daily food price data in this study shows weekly and annual seasonal patterns, making it suitable for modeling using the Seasonal Autoregressive Integrated Moving Average (SARIMA) method. This study aims to develop a web-based food price prediction system that provides daily price information and predictions for the next few days. The data used is daily data from January 2022 to October 2025 from the Bengkalis Regency Food Security Agency. The data was processed through cleaning, stationarity testing, model training, and prediction using Python with the statsmodels library, while the system was developed using the Flask framework and MySQL database. The results show that the SARIMA model is capable of producing predictions that follow historical trends and seasonal patterns. Evaluation using Root Mean Square Error (RMSE) shows that Minyakita has an RMSE of 1.415.04 or around 8–9% of the average price, while commodities with high fluctuations such as cayenne pepper and imported soybeans have higher error rates. The system's security was tested using OWASP ZAP.