Soft Computing and Machine Learning Approaches for Mapping in an Indian Coastal City for Flood Vulnerability

Flood susceptibility mapping is vital in mitigating flood risks and managing disaster preparedness. This research article proposes a hybrid machine learning model that combines Support Vector Machines (SVM) with Genetic Algorithm (GA) metaheuristic algorithm for making of a Flood Susceptibility Map (FSM) for Chennai city. The objective is to enhance the performance of machine learning in predicting flood-prone areas. Future research directions are suggested, and an analysis is done on the use of such machine learning models in mapping and assessing flood risk. Based on fifteen Flood Conditioning Factors (FCFs), the flood susceptibility model was built. The data from the flood inventory is then split into two categories: 70% is used to train the model and 30% is used to validate it. The accuracy of the generated model is obtained as 80%.