Optimization of Support Vector Regression using Genetic Algorithm and Particle Swarm Optimization for Rainfall Prediction in Dry Season

Abstract

Support Vector Regression (SVR) is Support Vector Machine (SVM) is used for regression case. Regression method is one of prediction season method has been commonly used. SVR process requires kernel functions to transform the non-linear inputs into a high dimensional feature space. This research was conducted to predict rainfall in the dry season at 15 weather stations in Indramayu district. The basic method used in this study was Support Vector Regression (SVR) optimized by a hybrid algorithm GAPSO (Genetic Algorithm and Particle Swarm Optimization). SVR models created using Radial Basis Function (RBF) kernel. This hybrid technique incorporates concepts from GA and PSO and creates individuals new generation not only by crossover and mutation operation in GA, but also through the process of PSO. Predictors used were Indian Ocean Dipole (IOD) and NINO3.4 Sea Surface Temperature Anomaly (SSTA) data. This research obtained an SVR model with the highest correlation coefficient of 0.87 and NRMSE error value of 11.53 at Bulak station. Cikedung station has the lowest NMRSE error value of 0.78 and the correlation coefficient of 9.01.