Bayesian Hierarchical Probabilistic Forecasting of Intraday Electricity Prices in Germany

Core Concepts

Bayesian forecasting of electricity prices in the German continuous intraday market incorporates parameter uncertainty for improved forecasting performance.

Abstract

The article introduces Bayesian forecasting of electricity prices in the German continuous intraday market, focusing on parameter uncertainty. It challenges the gold standard of using LASSO for feature selection and presents Orthogonal Matching Pursuit as a better alternative. The study highlights the importance of probabilistic forecasts for risk management in the context of renewable energy sources. The research contributes to the field of Electricity Price Forecasting by incorporating uncertainties in model parameters.

Stats

"In 2022, 134.6 TWh of the total of 611.21 TWh traded on EPEX was traded on the intraday market, a new all-time high." "The IDFull is the VWAP of all transactions of a product." "The average slope of quarter-hourly SDAC market clearance prices is used as a measure of elasticity."

Quotes

"According to the weak-form efficiency hypothesis, it would not be possible to significantly improve this benchmark built from last price information." "We challenge the declared gold standard of using LASSO for feature selection in electricity price forecasting."

Key Insights Distilled From

Bayesian Hierarchical Probabilistic Forecasting of Intraday Electricity Prices

by Dani... at arxiv.org 03-11-2024

https://arxiv.org/pdf/2403.05441.pdf

Deeper Inquiries

질문 1

모델 파라미터의 불확실성을 모델에 포함시키는 것이 전기 가격 예측의 정확도에 어떤 영향을 미치나요? Answer 1 here

질문 2

전기 가격 예측에서 LASSO 대신 Orthogonal Matching Pursuit를 사용하는 것이 특징 선택에 미치는 영향은 무엇인가요? Answer 2 here

질문 3

이 연구 결과를 전기 가격 예측 이외의 다른 시장이나 산업에 어떻게 적용할 수 있을까요? Answer 3 here