By analyzing and comparing the characteristics of neural network models and complex systems and the requirements of their modeling methodology, it is shown that the applications of the neural network in the complex system modeling are feasible. The main application fields of the neural network in the complex system modeling are discussed, such as the natural language processing, the speech recognition, the visual processing, and the fault diagnosis; and the main problems are pointed out, including the facts that the neural network can not describe the structure and the evolution of the system, it can not handle multiple tasks at the same time, and it can not solve the non repetitive problems well. Also the solutions of these problems are discussed preliminarily.

neural network  /  complex system  /  modeling