A Note on Averaging Day-Ahead Electricity Price Forecasts Across Calibration Windows

We propose a novel concept in energy forecasting and show that averaging day-ahead electricity price forecasts of a predictive model across 28-728 day calibration windows yields better results than selecting only one "optimal" window length. Even more significant accuracy gains can be achi...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on sustainable energy 2019-01, Vol.10 (1), p.321-323
Main Authors: Hubicka, Katarzyna, Marcjasz, Grzegorz, Weron, Rafal
Format: Article
Language:English
Subjects:
autoregression
Calibration
calibration window
combining forecasts
committee machine
Computational modeling
Diebold–Mariano test
Electricity
Electricity price forecasting
Electricity pricing
Forecasting
Load modeling
Mathematical model
Microsoft Windows
NARX neural network
Predictive models
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We propose a novel concept in energy forecasting and show that averaging day-ahead electricity price forecasts of a predictive model across 28-728 day calibration windows yields better results than selecting only one "optimal" window length. Even more significant accuracy gains can be achieved by averaging over a few, carefully selected windows.
ISSN:1949-3029
1949-3037
DOI:10.1109/TSTE.2018.2869557