Prediction of Rainfall in Sylhet Using Machine Learning Approach

Rainfall significantly impacts lots of areas of human life, Including natural calamities for example landslides, floods, droughts, agriculture, etc. Floods are one of the most serious disasters especially for the rural communities of our country as Sylhet. The recent flood caused a lot of damage to Sylhet in 2022, including the destruction of safe housing and healthy food as well as numerous mortality from a need for adequate medical care. Accurate rainfall forecasts can be beneficial for agriculture, aviation, natural disasters, floods, construction, and transportation industries. This study's objective is to build the prediction and analysis of probability. Machine learning is one of the most popular techniques used for rain prediction and probability. Total 21390 day data in Sylhet Weather Forecast. Weather Forecasting Dataset collected from Bangladesh Rice Research Institute website. The dataset contains Climate data recorded on a daily basis from 1956 to 2014. We applied data preprocessing use with label encoders, feature scaling and standard scalars to improve data behavior and performance, we removed null values and extraneous items. With the preprocessed data. we have run some classification algorithms to predict Rain or no rain, rain in millimeters, and Rain Category. Among them, we got the best accuracy results from the Random Forest Algorithm and KNN. In the next, we implemented the formula of Naive Bayes probability to find Rain or no rain and Rain Categories probability.