A hybrid model for GEFCom2014 probabilistic electricity price forecasting

Abstract

This paper provides detailed information on Team Poland’s winning methodology in the electricity price forecasting track of GEFCom2014. A new hybrid model extending the Quantile Regression Averaging (QRA) approach of Nowotarski and Weron (2015) is proposed. It consists of four major blocks: point forecasting, pre-filtering, quantile regression modeling and post-processing. This universal model structure enables a single block to be developed independently, without the performances of the remaining blocks being affected. The four-block model design is complemented by the inclusion of expert judgement, which may be of great importance in periods of unusually high or low electricity demand.