Predicting the Future: Using AdaBoost to Forecast Food Commodity Prices Across Indonesian Markets

Rice prices play a crucial role in Indonesia's food security, making accurate predictions essential for effective decision-making in price stabilization and policy planning. This study aims to predict rice prices in various markets across Indonesia by applying the Adaptive Boosting (AdaBoost) algorithm. Eight modeling schemes were explored using a combination of AdaBoost and Decision Tree Regression, with coefficient of determination <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\mathrm{R}^{2})$</tex> and Mean Absolute Percentage Error (MAPE) as evaluation metrics. The dataset, comprising 35,107 data points, was obtained from the Strategic Food Price Information Center (PIHPS) and covers the period from May 2023 to April 2024. The results indicate that Scheme 5, utilizing a 90:10 data split scenario, achieved the best performance with an <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{R}^{2}$</tex> value of 99.72% and a MAPE of 0.14%. This significantly outperforms previous research employing the Gradient Boosted Trees Regression algorithm, which achieved an <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathrm{R}^{2}$</tex> of 90.47% and a MAPE of 12.41%. These findings demonstrate the superiority of the AdaBoost algorithm in modeling food commodity prices, particularly in forecasting rice prices in Indonesia. This study contributes to the advancement of more accurate price prediction methods, which can be leveraged by policymakers, market stakeholders, and the agricultural sector to support price stabilization and food security management in Indonesia.