An approach for Flood Severity assessing Flood Impact Prediction using Hybrid Machine Learning Models

One major natural disaster that happens all around the world as a result of natural forces is flooding. They have resulted in animal losses, significant property damage and even fatalities. It is crucial to have a flood risk prediction system that is both accurate and effective. Early warning system helps to minimize any harm. It is crucial for protecting people and property by sending out timely alerts. This study employs and evaluates four different machine learning algorithms: Decision Tree Classifier, Random Forest Classifier, Logistic Regression and K-Nearest Neighbors (KNN) Classifier. Additionally, it includes a comprehensive analysis aimed at assessing flood susceptibility across the targeted region. Evaluation of these models' predictive power for flood susceptibility is the goal. The algorithms' classification accuracies for the provided datasets are 62%, 87%, 83% and 83% respectively. This study underscores the role of machine learning approaches in disaster management. It specifically delves into scholarly work on hazard prediction, disaster detection, early warning systems, monitoring, risk and vulnerability assessment, damage appraisal, post-disaster recovery and pertinent case studies.