Modeling mechanism of grey neural network and its application

In this paper, based on amalgamation of grey DGM(2,1) model and forward neural network, the grey neural network(DGNNM(2,l)) is put forward. Moreover, the relationship between parameters and the neural networks´ weights is presented, and the learning algorithm of the DGNNM(2,1) is also proposed. Because the learning error of the DGNNM(2,1) can be controlled effectively, the model´s precision is to be improved correspondingly. Finally, the grey neural network is used to build the model of Chinese per-capita building areas. The result of its application shows the effectiveness of the DGNNM(2,1).