Predicting Rainfall Using Machine Learning

Our project covers a deep dive analysis into weather forecasting and particularly determining factors that significantly affect rainfall. For our analysis we focus on the country of Australia, taking a dataset consisting of information regarding weather trends across different locations within the country for a span of ten years (2007-2017). We first perform an exploratory data analysis on our dataset, determining factors that are strongly correlated to each other.

In order to gain better clarity on the weather conditions across the country, we perform a time series analysis across the various locations within Australia, determining how the weather patterns have varied throughout the years. We then go on to identifying the parameters that most significantly affect rainfall. On obtaining these factors, we train a logistic and SVM model to determine if our model can predict whether rainfall would occur or not based on these significant parameters. We go on to compare the two models to see which performs better based on several parameters. We finally conclude our project based on future scopes, implementing counterfactuals to examine the impact of global warming on the country.

Github Link: https://github.com/prathameshmahankal/Predicting-Rainfall-Using-Machine-Learning