Performance Evaluation of hybrid ANFIS model for Flood Prediction
Dillip K. Ghose, Kshirabdhi Tanaya, Abinash Sahoo, Upendra Kumar
Abstract
Because of the catastrophic socio-economic effects of flood and the increase in its anticipated occurrence shortly, flood prediction has gained importance worldwide. Hence, a more consistent hydrologic prediction model is vital to plan, design, and manage water resources activities. This research introduces a hybrid model integrating an adaptive neuro-fuzzy inference system with harris hawks optimisation (ANFIS-HHO) for forecasting river flood events in Barak river basin, India. The robust model's performance is compared with simple ANFIS based on Nash-Sutcliffe efficiency (NSE), and root mean square error (RMSE). Results obtained indicate that the hybrid ANFIS-HHO model gave best performance with NSE of 0.9885, and RMSE of 61.87, implying its potential in flood forecasting. Also, the HHO algorithm can improve reliability of the standalone ANFIS model in flood forecasting and can solve problems of overfitting and underfitting during training of ANFIS model.