Comparable Investigation for Rainfall Forecasting using Different Data Mining Approaches in Sulaymaniyah City in Iraq

Abstract

Weather prediction is a critical assumption in weather forecasting. Weather prediction and has been one of the major scientifically and technologically demanding issues worldwide in the last century. The most significant parameter in a hydrological model is Rainfall. The meticulous Rainfall forecasting is one of the major demanding in the atmospheric research. The factors such as pressure, temperature, humidity, wind speed, mean sea-level etc. are used for rainfall forecasting. This study evaluates multiple classifiers such as Artificial Neural Network (ANN), Naïve Bayes and Support Vector Machine for rainfall prediction in Sulaymaniyah city and describes which one is most suitable to predict the precipitation. The dataset has been collected from weather forecast department in Sulaymaniyah city. Pre-processing technique such as cleaning and normalization processes is used for effective prediction. The data mining approaches are evaluated and the Performance is analyzed regarding precision, recall and f-measure with numerous ratios of training and test data.