Rainfall Classification using Support Vector Machine

Abstract

The quantity of rainfall that is likely to occur has a strong dependency on two very significant parameters namely humidity and temperature. The challenges involved in prediction of rainfall level have been the motivation behind writing this research paper. In this paper, SVM machine learning algorithm has been used to classify the input data, containing maximum temperature and humidity values, into two classes namely 'heavy rainfall' and 'light rainfall'. Two different SVM models are developed using linear and RBF kernels respectively using MATLAB. Finally, their classification performance is evaluated and compared.