Najnowsze trendy w prognozowaniu cen energii elektrycznej: uśrednianie prognoz 
po oknach kalibracji
[Recent advances in electricity price forecasting: averaging forecasts across 
calibration windows]

Tomasz Serafin, Rafał Weron
(Matematyka Stosowana, Wydział Matematyki, PWr, Wrocław; 
Katedra Badań Operacyjnych, Finansów i Zastosowań Informatyki, PWr, Wrocław)


A variety of methods and ideas have been tried for electricity price forecasting 
over the last two decades, with varying degrees of success. In the first part of 
the talk, Rafał Weron will provide a short overview of the recent advances. 
Then, in the second part, Tomasz Serafin will introduce a novel concept in energy 
forecasting and show that averaging day-ahead electricity price forecasts of a 
predictive model across calibration windows ranging from 28 to 1095 days yields 
better results than selecting only one "optimal" window length, not only 
ex-ante, but usually also ex-post. Even more significant accuracy gains can be 
achieved by averaging over a few carefully selected windows and assigning larger 
weights to windows that have performed well in the past.

Based on:
[1] K. Hubicka, G. Marcjasz, R. Weron (2018) A note on averaging day-ahead 
electricity price forecasts across calibration windows, IEEE Transactions on 
Sustainable Energy (https://doi.org/10.1109/TSTE.2018.2869557). Working paper 
version available from RePEc: https://ideas.repec.org/p/wuu/wpaper/hsc1803.html
[2] G. Marcjasz, T. Serafin, R. Weron (2018) Selection of calibration windows 
for day-ahead electricity price forecasting, Energies 11(9), 2364 
(https://doi.org/10.3390/en11092364).