Comparative Analysis of Neuro-Fuzzy and Support Vector Approaches for Flood Forecasting: Case Study of Godavari Basin, India

Forecasting and prediction have been a significant area of study for researchers since very past. Out of various approaches, soft computing data-driven models are very helpful for the purpose of forecasting. Soft computing models are usefully applicable when the relationship between the parameters is very complex to understand. India is a disaster-prone country which requires such major soft computing-based data-driven models to handle disasters like flood, drought and landslide. Flood has a major impact in many regions of India out of which Cauvery, Godavari and Ganges river basins are the most affected regions. The paper attempts to forecast floods by modeling river flow into the area of Godavari river basin of India which has a complicated topography. In this study, two data-driven techniques, adaptive neuro-fuzzy inference system (ANFIS) and support vector machine (SVM), were explored for the purpose of forecasting floods by predicting river flow in Cauvery river sub-basin of southern India.