PRediction of self-similar time series containing polynomial trends
Анатолій Васильович Омельченко, E. A. Rozdimakha, Олексій Валерійович Федоров · 2013
The analysis of prediction methods for time series that contain deterministic and stochastic trends is done. The method to extract a deterministic trend at presence of correlated noise is considered. Estimations for prediction accuracy depending on a prediction step for different algorithms are obtained by modeling.