Neural Network training model for weather forecasting using Fireworks Algorithm

Weather forecasting is the application of science and technology in order to predict the weather conditions. It is important for agricultural and industrial sectors. Models of Artificial Neural Networks with supervised learning paradigm are suitable for weather forecasting in complexity atmosphere. Training algorithm is required for providing weight and bias values to the model. This research proposed a weather forecasting method using Artificial Neural Networks trained by Fireworks Algorithm. Fireworks Algorithm is a recently developed Swarm Intelligence Algorithm for optimization. The main objective of the method is to predict daily mean temperature based on various measured parameters gained from the Meteorological Station, located in Bangkok. The experimental results indicate that the proposed method is advantageous for weather forecasting.