Natural disasters affect our nation almost every year. In particular for our nation’s rural communities like Sylhet and Sunamganj, the flood is one of them and the most deadly disaster. A lot of harm was done to the Sunamganj district in 2022 as a result of the recent flood, including removal of shelter and hygienic food, as well as many fatalities from lack of sufficient medical care. The ability to analyze data using methods like data mining and machine learning is becoming more and more accessible as technology advances. We may process our dataset using this strategy, generate predictions, and perform data analysis. The goal of this study is to forecast flood risk and flood depth in order to determine whether a flood will occur or not within the next few years. The goal of this study is to forecast flood risk and flood depth in order to determine whether a flood will occur within the next few years. Based on the national weather database, we created a sample dataset. The study also demonstrates different supervised machine learning techniques that we used on our dataset to forecast flood danger warning. We implemented data preprocessing techniques including feature scaling, Label encoders, and standard scalers to make the data behave and perform better when applying the models. In order to provide a better performance, we examined our dataset and removed any extraneous items. Then, we put our algorithms to work, and the results show that, when compared to other models, only the KNN classifier provides the best prediction accuracy (63.33%).