Rainfall prediction with neural network method and support vector regression

Weather is an important part of the daily life of living things, especially humans. Therefore humans need accurate, complete and fast weather information. Accurate weather prediction can be used to solve problems arising from weather effects such as drought detection, bad weather, plants and production, aviation, communication, and others. Scientists are now aware that data mining can be used as a tool for weather prediction. Human involvement is still needed to choose the best forecast model to be used as the basis for predictions. Weather prediction is an important application in meteorology and has become one of the most scientifically and technologically challenging problems in the whole world in the last century. The method used in this research is Neural Network and SVR (Support Vector Regression) which is the application of SVM (Support Vector Machine) for regression. The data used was obtained from Ngurah Rai Station, Bali Province. Some testing is done to get accurate prediction results. The results showed that the Neural Network method had a MAPE value of 0.85% while the SVR method had a MAPE value of 1.17%. It can be concluded that the Neural Network method has a smaller error value compared to the SVR method.