Most real-world time series have some degree of nonstationarity due to external perturbations of the observed system; external driving forces are the essential reason that leads to the nonstationarity of dynamics system. In this paper, the authors present a novel technique in which the authors incorporate external forces to predict nonstationary time series. To test the effect, the authors also examined two prediction experiments with an ideal time series from a logistic map and a proxy climate dataset for the past millennium. The preliminary results show that the resulting algorithm has better predictive ability than the one that does not consider the external forces.
In this paper, we investigate a novel technique that reconstructs the observed time series and incorporates driving forces. Furthermore, to illustrate and test the technique, we consider a couple of predictive experiments using ideal time series provided by the logistic and Lorenz systems with specific driving forces. The preliminary results show this approach can improve prediction proficiency to some extent, and the external forces play a similar role to that of state variables.
