Probabilistic predictions as inputs to statistical learning models: Price forecasting and decision support in markets for electricity
Grant no.: NCN 2023/49/N/HS4/02741
Funding agency: National Science Centre (NCN), Poland
Funding scheme: PRELUDIUM
Funding period: 1.02.2024-31.01.2027 (36 months)
Budget: 113 460 PLN
Title in Polish: Prognozy probabilistyczne jako dane wejściowe do modeli uczenia statystycznego: Prognozowanie cen i wspomaganie decyzji na rynku energii elektrycznej
Research team:
Principal Investigator (Kierownik):
Research Supervisor (Opiekun naukowy):
Collaborators (Współpracownicy):
– Ph.D. / M.Sc. / B.Sc. student
Aims and scope:
The energy landscape has become increasingly complex in recent years as renewable energy sources are constantly being expanded and integrated into the existing infrastructure to meet the European Union’s goal of carbon neutrality by 2050. At the same time, market participants require robust tools not only for predicting tomorrow’s prices, but also for quantifying the uncertain future. Hence, this project aims to develop novel approaches to electricity price forecasting (EPF) by integrating probabilistic predictions of fundamental variables as inputs to statistical learning models for wholesale electricity prices. The ultimate goal is to enable informed decision support. To achieve this, the project offers an integrated approach consisting of three interrelated and performed in parallel tasks:
Tasks:
- Developing and evaluating statistical learning models with probabilistic inputs to compute point predictions of electricity prices.
- Employing quantile forecasts of explanatory variables as inputs to statistical learning models for generating predictive distributions of electricity prices.
- Utilizing probabilistic predictions for decision support and economic evaluation of price forecasts.
Publications:
Peer-reviewed articles in JCR-listed journals
2024 (1+)
- A. Lipiecki, B. Uniejewski, R. Weron (2024) Postprocessing of point predictions for probabilistic forecasting of day-ahead electricity prices: The benefits of using isotonic distributional regression, Energy Economics 139, 107934 (doi: 10.1016/j.eneco.2024.107934). Julia codes available from GitHub
Forthcoming publications, submitted papers and work in progress
- B. Uniejewski, F. Ziel (2024) Probabilistic Forecasts of load and RES for Electricity Price Forecasting, work in progress.
- K. Chęć, B. Uniejewski, R. Weron (2024) Extrapolating the long-term seasonal component of electricity prices for forecasting in the day-ahead market, submitted. Working paper version available from RePEc: ideas.repec.org/p/ahh/wpaper/worms2404.html