Novel Hybrid Models by Coupling Support Vector Regression (SVR) With Meta-heuristic Algorithms (WOA and GWO) for Urban Flood Risk Mapping

Abstract Schools as social bases and children’s centers are among the most vulnerable areas to flooding. Flood risk mapping (FRM) is very important for flood preparedness and adopting preventive plans for reducing the school vulnerability to flooding. To achieve this, there is a need for the models that can be used in vast areas with high predictive accuracy. This study aims to develop the innovative hybrid models by coupling the support vector regression (SVR), statistical approaches, and two meta-heuristic algorithms, whale optimization algorithms (WOA) as well as grey wolf optimizer (GWO). According to the proposed methodology, a hybrid feature of SVR and frequency ratio (FR-SVR) is optimized by applying the GWO and WOA optimization algorithms to generate the maps related to flood susceptibility (FSMs). The method was utilized for the Ardabil Province located in southwestern Caspian Sea precincts of which faced devastating floods. The GIS database including 147 ground control locations of flooded zones and nine factors which influence flood were utilized to learn and ascertain the validity of the models. The statistical measures of RMSE, MAE, AUC, and ROC curve were then applied for the developed models in order to estimate prophetically. The results indicated that the meta-optimized FR-SVR-GWO as well as FR-SVR-WOA models exceeded the FR-SVR and FR models in training (RMSE FR-SVR-WOA = 0.2016, RMSE FR-SVR-GWO = 0.1885, AUC FR-SVR-WOA = 0.87, AUC FR-SVR-GWO = 0.88) and validation (RMSE FR-SVR-WOA = 0.2025, RMSE FR-SVR-GWO = 0.1986, AUC FR-SVR-WOA = 0.87, AUC FR-SVR-GWO = 0.87) phases. The FR-SVR-WOA and FR-SVR-GWO models were very competitive regarding AUC and RMSE values, but the FR-SVR-WOA model reproduced greater flood susceptibility rates and was considered for school flood risk mapping (SFRM).