Five-Class Flood Susceptibility Mapping Dataset with Imbalanced Labels

This dataset contains spatial environmental variables and corresponding five-class flood susceptibility labels (ranging from 0 to 4). It is designed for use in machine learning-based flood risk modeling and explainable artificial intelligence (XAI) applications. The dataset is imbalanced, with the majority of instances representing lower-risk categories and significantly fewer samples in higher-risk classes. The associated source code is publicly available at: https://github.com/shuimuxuanwo/Flood_Susceptibility_Modeling.