A Model-Free Method for Structual Change Detection Multivariate Nonlinear Time Series

• 出 处: Journal of Systems Engineering and Electronics
• 发表日期: 2003年02期
• 关键词: crossover otherwise Hefei distinguished empty Cambridge steps asymptotic breaks

In this paper, we apply the recursive genetic programming (RGP) approach to the cognition of a system, and then proceed to the detecting procedure for structural changes in the system whose components are of long memory. This approach is adaptive and model-free, which can simulate the individual activities of the system’s participants, therefore, it has strong ability to recognize the operating mechanism of the system. Based on the previous cognition about the system, a testing statistic is developed for the detection of structural changes in the system. Furthermore, an example is presented to illustrate the validity and practical value of the proposed.

In this paper, we apply the recursive genetic programming (RGP) approach to the cognition of a system, and then proceed to the detecting procedure for structural changes in the system whose components are of long memory. This approach is adaptive and model-free, which can simulate the individual activities of the system’s participants, therefore, it has strong ability to recognize the operating mechanism of the system. Furthermore, an example is presented to illustrate the validity and practical value of the proposed.