The Novel Modelling Method of Time Series Based on the Internal Model Principle
Wenbo Wang · Advances in intelligent systems research/Advances in Intelligent Systems Research · 2015
In this paper, a novel modelling method for time series is considered.Because the influences of the internal or external factors are often inevitable which result in the irregular change of the series, especially in many of the financial time series, the conventional modelling methods, which depend on the autocorrelation of the series, cannot meet the requirements or have a passive response to the factors.To find an effective method which can reflect the intrinsic essence or external influences of the series, the internal model principle, which is used in the synthesis of control systems for revealing the fundamental features of the factors, is introduced.It may open up broad prospects for improving the fitness or prediction ability of the models under the circumstances mentioned above.This paper discusses the possibility and provides some analytical results.