Structure change analysis of complex system
Qing Sun · Journal of systems engineering · 2000
Traditional variable structure analysis depends on models. At present, structural change analysis focuses on the parameter's changes and variable's changes in quantitative models. These traditional analysis based on condition of researchers have acknowledge about the system rules and can provide the model forms. But in the actual systems, it is difficult to know the system rule before data analysis process, the traditional analysis method can not be suitable to complex systems. In order to solve this problem, we propose a new idea to detecting structure changes in complex system. We apply neural network technique to analyze system running rules, establish two kinds of algorithms for detecting structural change in actual system. As an empirical case, we discuss the Shanghai Stock Market using our new algorithm, and have got satisfied result.