An Efficient Weather Forecasting System using a Hybrid Neural Network SOFM–MLP

Abstract

Weather prediction is a challenging task for researchers and has drawn a lot of research interest in the recent years. Literature studies
have shown that machine learning techniques achieved better performance than traditional statistical methods. Presently multilayer perceptron
networks (MLPs) are used for prediction of the maximum and the minimum temperatures based on past observations on various atmospheric
parameters. To capture the seasonality of atmospheric data, with a view to improving the prediction accuracy, a novel weather forecasting
system is presented in this paper. The proposed system is based on a neural architecture that combines a selforganizing feature map (SOFM) and
MLPs to realize a hybrid network named SOFM–MLP. It is also demonstrated that the use of appropriate features such as temperature gradient
can not only reduce the number of features drastically, but also can improve the prediction accuracy. These observations motivated us to use a
feature selection MLP (FSMLP) instead of MLP, which can select good features online while learning the prediction task. FSMLP is used as a
preprocessor to select good features. The combined use of FSMLP and SOFM–MLP provides better result in a network system that uses only
very few inputs but can produce good prediction. The proposed system is experimented using the real time data observations and from which it
is found that the proposed system predict the temperature with minimum error.

 

Keywords: Atmospheric science, back propagation, feature selection, neural networks, self-organizing feature map (SOFM), temperature
forecasting