arXiv：cs.LG· Hao Wang, Licheng Pan, Yuan Lu, Zhixuan Chu, Xiaoxi Li, Shuting He, Zhichao Chen, Haoxuan Li, Qingsong Wen, Zhouchen Lin·· 7 小时前AI 评分40
DistDF：时间序列预测需要联合分布 Wasserstein 对齐
DistDF: Time-Series Forecasting Needs Joint-Distribution Wasserstein Alignment
AI 导读
DistDF 通过最小化预测序列与标签序列条件分布之间的分布差异来实现对齐，并提出可证明上界该条件差异的联合分布 Wasserstein 散度，该散度可微且兼容梯度优化。实验显示 DistDF 能提升多种预测模型并达到领先性能，代码已开源，论文已被 ICLR 2026 接收。
正文
Abstract:Training time-series forecasting models requires aligning the conditional distribution of model forecasts with that of the label sequence. The standard direct forecast (DF) approach resorts to minimizing the conditional negative log-likelihood, typically estimated by the mean squared error. However, this estimation proves biased when the label sequence exhibits autocorrelation. In this paper, we propose DistDF, which achieves alignment by minimizing a distributional discrepancy between the conditional distributions of forecast and label sequences. Since such conditional discrepancies are difficult to estimate from finite time-series observations, we introduce a joint-distribution Wasserstein discrepancy for time-series forecasting, which provably upper bounds the conditional discrepancy of interest. The proposed discrepancy is tractable, differentiable, and readily compatible with gradient-based optimization. Extensive experiments show that DistDF improves diverse forecasting models and achieves leading performance. Code is available at this https URL.
| Comments: | Accepted by ICLR 2026 |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2510.24574 [cs.LG] |
| (or arXiv:2510.24574v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2510.24574 arXiv-issued DOI via DataCite |
|
| Journal reference: | ICLR 2026 |
Submission history
From: Hao Wang [view email]
[v1]
Tue, 28 Oct 2025 16:09:59 UTC (1,557 KB)
[v2]
Sat, 11 Apr 2026 06:43:04 UTC (1,678 KB)
[v3]
Tue, 6 Oct 2026 12:10:37 UTC (1,711 KB)
来源：arXiv：cs.LG · arxiv.org