nabqr: Python package for improving probabilistic forecasts

2025-01-29

AI Summary

To address the insufficient accuracy—and particularly the low reliability—of day-ahead probabilistic wind power forecasting in Denmark, this paper introduces NABQR, an open-source Python toolkit. Methodologically, it proposes the first end-to-end, scenario-level calibration framework integrating neural adaptive basis functions with time-adaptive quantile regression: an LSTM models the dynamic characteristics of forecast errors to drive time-aware optimization of quantile weights, while incorporating probabilistic calibration and multi-scenario post-processing. Empirical evaluation on Danish onshore and offshore wind farms demonstrates up to a 40% reduction in mean absolute error, significantly improving both the sharpness and calibration of quantile forecasts—and thereby enhancing operational usability. The core contribution is the first neural adaptive calibration paradigm enabling time-dynamic quantile adjustment.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Relational Probabilistic Models
• Game Theory and Economic Paradigms: Adversarial Learning

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

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.

Problem

• Wind Power Forecasting
• Denmark
• Day-Ahead Prediction

Innovation

• LSTM Network
• Time Adaptive Quantile Regression
• Open-Source Python Package