Machine Learning Approach to Predict Rainfall Amount of Dhaka

Abstract

Weather Prediction is getting very popular in the current era of Artificial Intelligence. And it is very important to predict rainfall amount in a country like Bangladesh where 50% of people is a farmer. Because the farming in Bangladesh is based on Natural Monsoon. Farmers need to have an idea of how much rain may fall in the coming days. Also, rainfall information helps people become aware of several potential natural disasters. Rainfall prediction is getting very important for Bangladesh's economy day by day. Light or heavy - both rainfalls affect rural and urban life. This work aims to find out the pattern of the average rainfall of Dhaka per month. I used different machinelearning algorithms to predict the future rainfall amount of Dhaka such as Simple linear regression, Multivariate linear regression, Polynomial regression. Because rainfall amount depends on several weather attributes, in linear regression and polynomial regression, we used different attributes to find out which attribute gives the best result. And for multivariate linear regression, we used all possible attribute that has a relation with rainfall.