Weather Forecasting for Batu Pahat Using Neural Network

Nowadays, weather forecasting plays a vital role in human activities. The complexity of data gathering in the meteorology department and high technical costs lead to inaccurate weather forecasting. Due to minimizing this problem, an application of using an artificial neural network (ANN) has been developed to forecast weather conditions using the Matlab compiler. This application will help users to define the weather conditions daily well and made early preparation to encounter uncertainty. This application performs in Multilayer Perceptron (MLP) using the back-propagation (BP) algorithm. The data is undergoing training and testing based on actual data obtained from the Malaysian Meteorological Department. The result will show in graph plotting for the training and testing process. Predictive accuracy for each step of stimulation will measure using Mean Square Error (MSE) graph in this application. Therefore, this application will be able to assist the Malaysian Meteorological Department to forecast weather, so our government and people have sufficient time to prepare and solve.