nabqr: Python package for improving probabilistic forecasts

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Jørgensena, Bastian Schmidt, Møller, Jan Kloppenborg, Nystrup, Peter, Madsen, Henrik
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866912209340203008
author Jørgensena, Bastian Schmidt
Møller, Jan Kloppenborg
Nystrup, Peter
Madsen, Henrik
author_facet Jørgensena, Bastian Schmidt
Møller, Jan Kloppenborg
Nystrup, Peter
Madsen, Henrik
contents We introduce the open-source Python package NABQR: Neural Adaptive Basis for (time-adaptive) Quantile Regression that provides reliable probabilistic forecasts. NABQR corrects ensembles (scenarios) with LSTM networks and then applies time-adaptive quantile regression to the corrected ensembles to obtain improved and more reliable forecasts. With the suggested package, accuracy improvements of up to 40% in mean absolute terms can be achieved in day-ahead forecasting of onshore and offshore wind power production in Denmark.
format Preprint
id arxiv_https___arxiv_org_abs_2501_17604
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle nabqr: Python package for improving probabilistic forecasts
Jørgensena, Bastian Schmidt
Møller, Jan Kloppenborg
Nystrup, Peter
Madsen, Henrik
Machine Learning
Applications
Computation
We introduce the open-source Python package NABQR: Neural Adaptive Basis for (time-adaptive) Quantile Regression that provides reliable probabilistic forecasts. NABQR corrects ensembles (scenarios) with LSTM networks and then applies time-adaptive quantile regression to the corrected ensembles to obtain improved and more reliable forecasts. With the suggested package, accuracy improvements of up to 40% in mean absolute terms can be achieved in day-ahead forecasting of onshore and offshore wind power production in Denmark.
title nabqr: Python package for improving probabilistic forecasts
topic Machine Learning
Applications
Computation
url https://arxiv.org/abs/2501.17604