Analysis of Support Vector Machine (SVM) implemented on prediction of production rice commodity that can help the management of rice production in Indonesia. Prediction is done with Matlab R2016A especially function of SVM Regression. The prediction results were evaluated by performance criteria such as Root Mean Squared Error (RMSE), R-Squared and Adjusted R-Squared, and also curve fitting. SVM parameters determined automatically after processing is completed. Predictions done annually,

more »

... ed from 2006 to 2015. The results of those predictions determined the value of performance to get the value of the correspondence between the predicted value and the actual value and the best prediction is illustrated by curve fitting. It also conducted comparison performance of predictions per year to determine which ones produce the best fit. Results of prediction rice commodity with SVM method showed that the best fit is prediction in 2007 with RMSE value of 1.20E+06, R-Square of 0.794 or 79.4%, Adjusted R-Square of 0788 or 78.8%, as well as curve fitting shows the level distribution predictions are optimal for the year.