Skip to main content
QUICK REVIEW

[论文解读] CoRA: Boosting Time Series Foundation Models for Multivariate Forecasting through Correlation-aware Adapter

Hanyin Cheng, Xingjian Wu|arXiv (Cornell University)|Mar 23, 2026
Time Series Analysis and ForecastingComputer Science被引用 0
一句话总结

CoRA 是一个轻量级、即插即用的相关性感知适配器,用于时间序列基础模型,在微调过程中联合建模动态、异质和部分信道相关性,以提升多变量预测能力。

ABSTRACT

Most existing Time Series Foundation Models (TSFMs) use channel independent modeling and focus on capturing and generalizing temporal dependencies, while neglecting the correlations among channels or overlooking the different aspects of correlations. However, these correlations play a vital role in Multivariate time series forecasting. To address this, we propose a CoRrelation-aware Adapter (CoRA), a lightweight plug-and-play method that requires only fine-tuning with TSFMs and is able to capture different types of correlations, so as to improve forecast performance. Specifically, to reduce complexity, we innovatively decompose the correlation matrix into low-rank Time-Varying and Time-Invariant components. For the Time-Varying component, we further design learnable polynomials to learn dynamic correlations by capturing trends or periodic patterns. To learn positive and negative correlations that appear only among some channels, we introduce a novel dual contrastive learning method that identifies correlations through projection layers, regulated by a Heterogeneous-Partial contrastive loss during training, without introducing additional complexity in the inference stage. Extensive experiments on 10 real-world datasets demonstrate that CoRA can improve TSFMs in multivariate forecasting performance.

研究动机与目标

  • 通过利用TSFM常忽略的信道间相关性来提升多变量预测的动机。
  • 提出一个轻量级插件,可以与TSFM一起微调,而无需重新训练主干网络。
  • 在低推断复杂度下建模三种相关性类型(动态、异质、部分)。
  • 在真实世界数据集上验证有效性,并与现有相关插件进行比较。

提出的方法

  • 引入动态相关性估计(DCE),通过低秩Q_t与时不变V分解学习时变相关性。
  • 将可学习的相关性分解为 M_t^corr = R + Q_t V Q_t^T,结合基于规则的R与可学习分量。
  • 使用时序感知多项式建模演变的动态相关性,具有共享基底q和时变系数C_t。
  • 通过信道感知投影(HD)将信道划分为正相关空间和负相关空间,以捕捉异质相关性(HCorr)。
  • 应用非均匀相关性对比学习(H-PCorr Contrastive Learning),在双重对比损失引导下对强相关信道进行对齐、对弱相关信道进行分离。
  • 将正向与负向表示融合,并通过一个门控融合作用与原始TSFM预测合并,获得最终预测结果。
Figure 1: (a) Illustration of three different types of correlations, the formal definitions are provided in Appendix A . (b) Comparisons of different plugins for learning correlations
Figure 1: (a) Illustration of three different types of correlations, the formal definitions are provided in Appendix A . (b) Comparisons of different plugins for learning correlations

实验结果

研究问题

  • RQ1CoRA 能否在不重新训练骨干网络的前提下,提升现有TSFM在多变量时间序列上的预测性能?
  • RQ2相关性分解为动态部分与不变部分的效果如何,时间感知多项式在建模DCorr方面的影响?
  • RQ3HCorr 与 PCorr 是否需要专门的训练信号,对比学习能否在微调阶段引导其分离?
  • RQ4在不同信道数量的数据集上,CoRA 的推理时复杂度与实际效率表现如何?

主要发现

  • CoRA 在十个真实数据集的多变量设置中提升了TSFM的预测性能。
  • 提出的带有时变和时不变分量的DCorr分解在表达能力上可与加法分解相媲美,但参数成本更低。
  • 时序感知多项式提供了以可调阶数K建模动态相关性的机制,在精度与效率之间取得平衡。
  • 异质相关性划分加上H-PCorr对比学习实现了HCorr和PCorr的自适应学习,而不增加推理时复杂度。
  • 效率分析显示在微调和推理阶段CoRA的额外开销很小,在信道数量量级上保持线性复杂度。
  • 与其他相关插件的对比评估表明,针对TSFMs设计的CoRA在少量-shot微调场景下具有更优的性能。
Figure 2: The framework of CoRA. (a) CoRA begins by learning DCorr in Dynamic Correlation Estimation module. Heterogeneous Division module projects representations into positive and negative spaces for HCorr. Then CoRA conducts H-PCorr Contrastive Learning in each space to guide projection and captu
Figure 2: The framework of CoRA. (a) CoRA begins by learning DCorr in Dynamic Correlation Estimation module. Heterogeneous Division module projects representations into positive and negative spaces for HCorr. Then CoRA conducts H-PCorr Contrastive Learning in each space to guide projection and captu

阅读本论文

在 Nubint 阅读器中打开原文,边读边就任意段落提问。

在 Nubint 中阅读本论文

本解读由 AI 根据论文原文生成,引用前请核对原文。