arXiv：cs.LG· Xinyi Yi, Moy Yuan, Ioannis Lestas·· 3 小时前AI 评分35
TRACE：基于官方运营文本的可复现电价预测基准
TRACE: A Reproducible Benchmark for Electricity Price Forecasting with Official Operational Text
AI 导读
TRACE 是一个可复现的电价预测基准，包含 7,300 个 zone-day 实例，将美国某主要市场五个区域的电价与预测截止时点可获取的官方运营文本配对，并重建截止时点文本以防止未来信息泄漏。在时间序列基础模型上，该文本使上尾 pinball loss 中位数降低 7.4%；跨天文本错配消融实验则使增益反转，低于无文本基线。该工作已被 NeurIPS 2026 FMTS Workshop 接收。
正文
Abstract:Electricity price forecasting (EPF) supports scheduling, bidding, and risk management in electricity markets, yet existing benchmarks focus mainly on numerical inputs, leaving the forecasting value of forecast-time textual context insufficiently evaluated. We introduce TRACE, a reproducible benchmark of 7,300 zone--day instances pairing prices from five zones in a major U.S. market with official operational text available at the forecast cutoff. TRACE reconstructs official operational text at each cutoff, preventing post-cutoff information leakage. We evaluate TRACE for semantic alignment and forecasting value. Semantic assessments align with central movement and both tail risks in ground-truth prices, most consistently for upper-tail price risk. Forecasting value is reflected in a median 7.4\% reduction in upper-tail pinball loss across time-series foundation models. A controlled cross-day text-mismatch ablation reverses the gains, falling below the no-text baseline.
| Comments: | Accepted to the NeurIPS 2026 Workshop on Foundation Models for Time Series (FMTS) |
| Subjects: | Machine Learning (cs.LG); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.02256 [cs.LG] |
| (or arXiv:2610.02256v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02256 arXiv-issued DOI via DataCite |
Submission history
From: Xinyi Yi [view email]
[v1]
Wed, 30 Sep 2026 20:44:33 UTC (162 KB)
来源：arXiv：cs.LG · arxiv.org