Utilizing the Sarima Model and Support Vector Regression to Forecast Monthly Rainfall in Bandung City

Abstract

As one of the largest cities in Indonesia, Bandung has varying monthly rainfall intensity. High rainfall is very dangerous for people's lives and will have an impact on various sectors such as agriculture, fisheries, tourism, and transportation. For this reason, rainfall prediction is needed as an effort for the government to make policies and the community can anticipate the possibility of high rainfall that occurs. This study compares the effectiveness of SARIMA and Support Vector Regression (SVR) models in predicting monthly rainfall objectively, with the aim of improving decision making for stakeholders. Forecasting rainfall data is carried out based on the best method of the two methods that have been compared. The results showed that the SARIMA method outperformed the SVR method in forecasting precision, as seen from the lower RMSE value of 93.2045. The results provide valuable insights into weather prediction methodologies, benefiting authorities and the public.