Machine Learning-Based Flood Prediction of Assam

Hood prediction is crucial in disaster management and mitigation, particularly in flood-prone regions like Assam. This paper presents a study on machine learning-based flood prediction using Assam's historical rainfall and geospatial data. Among the five machine-learning algorithms assessed, the Run-dom Forest algorithm was the best-performing model for flood prediction. The proposed system can predict the risk of floods in the next live days by using forecast rainfall data obtained from the OpenWeather API, total rainfall data, and preprocessed geugraphical data. A user-friendly web application was also created to enhance Its usability and practicality. The proposed system can accurately predict floods, which will help disaster preparedness in Assam and other flood-prone regions worldwide.