The application of dynamic self-organised multilayer network inspired by the Immune Algorithm for weather signals forecast

Neural network architecture called Dynamic Self-organised Multilayer Network Inspired by the Immune Algorithm is proposed for the prediction of weather signals. Two sets of experiments have been implemented. The simulation results showed slight improvement achieved by the proposed network when using the average results of 30 simulations. For the second set of experiments, the simulation results indicated that there is no significant improvement over the first set of experiments. Since clustering methods have been widely used in different applications of data mining, the adaption of unsupervised learning in the proposed network might serve these different applications, for example, medical diagnostics and pattern recognition for big data. The structure of the proposed network can be modified for clustering tasks by changing the back-propagation algorithm in the output layer. This can extend the application of the proposed network to scientifically analyse different types of big data.