Comparative analysis of advanced machine learning classifiers based on feature engineering framework for weather prediction

Rainfall is considered one of the most significant phenomena in the weather system, and as such its rate is one of the most crucial variables. In order to develop a prediction model by standard approaches, meteorological experts attempt to detect the atmospheric attributes such as sunlight, temperature, humidity and cloudiness etc. Machine Learning (ML) techniques are recently more evolved which provides results that are more satisfactory than those of traditional methods and are simple to use. This paper presents the ML classifiers such as Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Light Gradient Boost Machine (LGBM), Cat Boost (CB), and Extreme Gradient Boost (XGB) to predict the rainfall of the next day using feature engineering framework. The Area under the Receiver Operating Characteristic (AUROC) curve and the other statistical indicators such as recall, accuracy, precision, and Cohen kappa are employed to predict and compare the success rate of the above mentioned approaches. The validation results of the models in terms of AUROC values are XGB (0.99) > CB (0.98) > LGBM (0.95) >RF (0.98) >DT (0.91) > LR (0.88). The outcome of the XGB model establishes that it is more appropriate for the prediction of weather.