Flood Prediction in Bangladesh Using Machine Learning and Explainable AI: A Comparative Study

Floods are among the most devastating natural disasters, posing significant threats to life, property, and economic stability, particularly in developing countries like Bangladesh. Accurate flood prediction remains a critical challenge, requiring the integration of advanced data analysis and machine learning techniques. This paper presents a comprehensive study using a dataset tailored specifically for Bangladesh to forecast floods. Five machine learning models—Logistic Regression, Random Forest, Gradient Boosting, XGBoost, and Long Short-Term Memory (LSTM) networks—were employed to predict flood occurrences. Among these, the Gradient Boosting model achieved the highest accuracy of 95.38%. To enhance the interpretability of the model’s predictions, Explainable AI techniques, specifically SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-agnostic Explanations), were applied, providing insights into the model’s decision-making process. This work not only contributes to the field of flood prediction in Bangladesh but also offers valuable information for policymakers to implement more effective flood management strategies.