Prediction of Rainfall Using Hybrid SVM-HHO Model

Abstract

Abstract Rainfall modeling is considered a need of the hour in order to understand, control, and monitor the quality and quantity of water resources. Modeling rainfall is one of the key components of the meteorological engineering process. Hydrological and climatological studies require accurate rainfall estimation in order to effectively manage water resources. Thus, adopting an advanced, reliable, and intelligent model for forecasting will be of great value in water resources engineering. The present study tried to establish a hybrid model with the combination of Support Vector Machine (SVM) and Harris Hawks Optimization (SVM-HHO) for predicting the rainfall time series of Cachar district located in Assam, India. The predictive performance of models is examined based on statistical analysis. Statistical measures like root mean squared error (RMSE) and coefficient of correlation (CC) is used to evaluate the considered hybrid model. It is observed from obtained results that proposed hybrid model exhibited least RMSE value of 20.29, and highest CC value of 0.9714, during testing period. Findings of this study confirm that proposed hybrid optimisation strategy can be regarded as a powerful forecasting tool for achieving better generalisation capability and higher forecasting accurateness.