Adaptive Problem Decomposition in Cooperative Coevolution of Elman Recurrent Networks for Time Series Prediction

Cooperative coevolution employs different problem decomposition methods to decompose the neural network training problem into subcomponents. The efficiency of a problem decomposition method is dependent on the neural network architecture and the nature of the training problem. The adaptation of problem decomposition methods has been recently proposed which showed that different problem decomposition methods are needed at different phases in the evolutionary process. This paper employs an adaptive cooperative coevolution problem decomposition framework for training Elman recurrent networks on chaotic time series problems. The Mackey Glass, Lorenz and Sunspot chaotic time series are used. The results show improvement in performance when compared to cooperative coevolution and other methods from literature.