Machine Learning to Predict Food Prices in Aceh Province Using the Fuzzy Time Series Method Based on Average

This study aims to develop a food commodity price prediction system based on Fuzzy Time Series (FTS) using average-based methods, with a case study of price data from 2018 to 2023. The system is designed to predict the prices of five main commodities: Super Quality Rice, Fresh Chicken Meat, Fresh Chicken Eggs, Bulk Cooking Oil, and Premium Quality Sugar. The prediction process involves constructing the Universe of Discourse, intervals, and fuzzy logic relations (FLR and FLRG) to model historical price patterns. The results show that this model provides accurate predictions, with the best Mean Absolute Percentage Error (MAPE) value of 0.49% for Super Quality Rice, while MAPE for other commodities ranges from 0.69% to 1.44%. The comparison graph between actual data and prediction results demonstrates consistent pattern alignment, suitable for commodities with both high price fluctuations and stable trends. This system proves effective in projecting future food prices with low error rates, making it a reliable tool to support strategic decision-making in managing food commodity prices during the five-year analysis period.