arXiv cs.AI· Junhyeok Kang, Sangjun Han, Jun Seo, Seunghan Lee, Jaehoon Lee, Dongwan Kang, Minjae Kim, Sungdong Yoo, Tae Yoon Lim, Hwanil Choi, Soonyoung Lee, Wonbin Ahn·· 1 天前AI 评分29
QUAL：分位数对齐提升零样本时间序列基础模型
QUAL: Quantile Alignment Improves Zero-Shot Time-Series Foundation Models
AI 导读
研究者提出 QUAL（Quantile alignment），一种无需标签的框架，通过模型池作为概率参考对成员预测做分位数对齐，再经 Huber 聚合与重排得到稳健组合预测，无需训练或修改模型。在 GIFT-Eval 上，分位数对齐使成员平均 MASE 和 CRPS 分别降低 7.4% 和 9.8%，且在候选池混入无法预先验证质量的弱预测器时仍保持稳健。代码已开源。
正文 · 原文
Authors:Junhyeok Kang, Sangjun Han, Jun Seo, Seunghan Lee, Jaehoon Lee, Dongwan Kang, Minjae Kim, Sungdong Yoo, Tae Yoon Lim, Hwanil Choi, Soonyoung Lee, Wonbin Ahn
Abstract:Time-series foundation models (TSFMs) provide zero-shot forecasts for unseen series, but their accuracy varies across datasets. Combining several models can reduce this variability, yet zero-shot deployment provides no target labels for selecting or weighting them. We find that a model pool is useful not only as a final ensemble, but also as a probability reference for transforming each member forecast. Aligning quantiles through this pool-relative reference improves every member forecast in aggregate. Motivated by this finding, we propose QUAL (Quantile alignment), a simple label-free framework for improving and combining diverse TSFMs. QUAL first produces aligned member forecasts, then applies Huber aggregation and rearrangement to obtain a robust combined forecast without training or model modification. On GIFT-Eval, quantile alignment reduces average member MASE and CRPS by 7.4% and 9.8%. The full framework remains robust when weak forecasters enter a candidate pool whose quality cannot be verified in advance. Code is available at this https URL.
| Comments: | Accepted at the NeurIPS 2026 Workshop on Foundation Models for Temporal Systems: From Forecasting to World Modeling |
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.41100 [cs.AI] |
| (or arXiv:2610.41100v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.41100 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Junhyeok Kang [view email]
[v1]
Fri, 30 Oct 2026 00:41:39 UTC (214 KB)
来源：arXiv cs.AI · arxiv.org