Improving forecasting by estimating time series structural components across multiple frequencies

Kourentzes, Nikos and Petropoulos, Fotios and Trapero, Juan Ramon (2014) Improving forecasting by estimating time series structural components across multiple frequencies. International Journal of Forecasting, 30 (2). pp. 291-302. ISSN 0169-2070

[thumbnail of Kourentzes et al 2013 Multiple Aggregation Prediction Algorithm]
PDF (Kourentzes et al 2013 Multiple Aggregation Prediction Algorithm)
Kourentzes_et_al_Multiple_Aggregation_Prediction_Algorithm.pdf - Submitted Version

Download (311kB)


Identifying the appropriate time series model to achieve good forecasting accuracy is a challenging task. We propose a novel algorithm that aims to mitigate the importance of model selection, while increasing accuracy. From the original time series, using temporal aggregation, multiple time series are constructed. These derivative series highlight different aspects of the original data, as temporal aggregation helps in strengthening or attenuating the signals of different time series components. In each series the appropriate exponential smoothing method is fitted and its respective time series components are forecasted. Subsequently, the time series components from each aggregation level are combined, and then used to construct the final forecast. This approach achieves better estimation of the different time series components, through temporal aggregation, and mitigates the importance of model selection through forecast combination. Empirical evaluation of the proposed framework demonstrates significant improvements in forecasting accuracy, especially for long-term forecasts.

Item Type:
Journal Article
Journal or Publication Title:
International Journal of Forecasting
Additional Information:
The final, definitive version of this article has been published in the Journal, International Journal of Forecasting 30 (2), 2014, © ELSEVIER.
Uncontrolled Keywords:
ID Code:
Deposited By:
Deposited On:
17 Sep 2013 08:12
Last Modified:
22 Sep 2023 00:21