Prediction of Rainfall in Bangladesh: A Case Study of the Machine Learning

The main goal of this paper is to predict rainfall using Regression model of Machine Learning over time series so that steps can be taken beforehand to prevent disasters.The goal of weather forecasting is to provide information to everyone so that they can limit losses and assure safety, which includes life and property safety, as well as human health and wellness.Human eyes can miscalculate a large quantity of data in weather forecasting, causing calculations to go awry. In this paper, a machine learning approach was utilized to enhance accuracy and precision of weather forecasting. The key goal is to create a successful prediction system that can forecast Bangladesh’s rainfall in the coming years. The prediction of rainfall are run on weather data of Bangladesh from 1901 to 2015. By training our system with the collected data and tuning Random Forest Regression, we were able to achieve a 91% accuracy. Following the completion of the architecture, our adjusted model is compared to the base model, Randomized Search CV, using few evaluation metrics such as Mean Absolute Error (MAE), Mean Squared Logarithmic Error (MSLE), Root Mean Squared Logarithmic Error (RMSLE),and the results are quite promising. We can utilize machine learning in field weather forecasting to save lives and safeguard assets.