Support Vector Regression Algorithm for Predicting Rice Production In West Java Province

Rice is an important staple food in Indonesia, and increasing rice production is essential to meet the growing demand for food due to population growth. This study uses the Support Vector Regression (SVR) algorithm to predict rice production in West Java Province by considering variables such as harvest area, population, inflation, food GRDP, temperature, rainfall, length of sunshine, and the Highest Retail Price (HET) of fertilizer. The result is that harvest area, population, food GRDP, temperature, rainfall, and sunshine duration positively correlate with rice production, while inflation and fertilizer price ceilings correlate negatively. The analysis also showed excellent performance, with a MAPE value of 5.60%, indicating a small percentage of prediction error and an R-squared of 0.8936, indicating that the SVR model can explain 89.36% of the variation in rice production. To support sustainable rice production, agricultural policies should strengthen subsidies and control fertilizer prices, manage inflation, develop climate-smart infrastructure, and utilize data for evidence-based decision-making.