Prediksi Awal Musim Hujan Menggunakan Data Southern Oscillation Index dengan Support Vector Regression
Abstract
Climate is one of the important aspects in human life. The information about prediction and forecasting is required, especially in agricultural sector. However, the current prediction for rainy season onset using linear regression method is not very good.The main objective of this research is to develop a model using support vector regression to predict the onset of the rainy season, in order to get the accuracy of climate information. Rainy season onset prediction has been attempted at 1 weather stations in Indramayu. The data used in this study is the Southern Oscillation Index (SOI) from June to August and the onset of the rainy season (AMH) from 1979-2008. The domain of the SOI was selected based on the correlation. Prediction result was evaluated using the root mean squared error and squared correlation coefficient. The average squared correlation coefficient value obtained was 0.7 and the root mean squared error was 2.3 using the RBF kernel function.
Collections
- UT - Computer Science [2254]