Prediction of Rainfall in North Sumatera Using Machine Learning

Due to the large number of variables involved, weather prediction is one field with a wealth of data but also a higher degree of difficulty. Probabilistic models are used to generate forecasts, which are often not particularly accurate because they have a margin of error. In this article, we report on an exploratory study that looked into using machine learning to predict the rainy season. We employed some of the most significant machine learning methods (multinomial NB, decision tree, random forest, logistic regression, and SGD) in conjunction with a set of data detailing rainfall observations gathered in Northern Sumatera's major cities during the previous ten years. The approach based on logistic regression is the best, according to the data.