Comparison of SARIMA and SVM model for rainfall forecasting in Bogor city, Indonesia

Heavy rainfall and flash floods are still significant for several cities in Indonesia, especially Bogor City. Rainfall prediction is essential to develop early warning systems for early detection of massive and flash floods. The accurate forecasting method is needed to produce accurate predictions of rainfall. Several methods have been developed to get a precise rainfall forecast, such as the classical approach, such as seasonal autoregressive integrated moving average (SARIMA), and machine learning approach, for example, gradient boosting and support vector machine (SVM). In this study, we introduce for forecasting rainfall at Bogor city, Indonesia. The SVM provides an accurate result with a small MAPE.