Analysis of Machine Learning Techniques for Weather Forecasting

Investigation of weather forecasting (WF) is a critical task in many applications. Accurate WF is an all-time demand of these applications. This research aims to validate various machine learning (ML) classifiers for the predictions of rainy weather. The Koggle weather dataset is used for the validation and testing of the various ML methods, such as support vector machine (SVM), decision tree (DT) and artificial neural network (ANN). In this research, extended surveys of ML and NN-based WF methodologies are also provided. The prediction accuracy is compared for the validation of the various ML-based classifiers. 4 features are used for classifying the 5 classes of data. Overall, it is found that Random Forest and ANN have a validated accuracy of 84.35% each.