arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2610.17864cs.LGcs.AI
收藏

TStruct:学习用于长期时间序列预测的共享时间结构

TStruct: Learning Shared Temporal Structures for Long-Term Time Series Forecasting

University of Science and Technology of China (USTC)(中国科学技术大学);Suzhou Institute for Advanced Research, USTC(中国科学技术大学苏州高等研究院)

Yangze Li, Zhengyang Zhou, Qihe Huang, Pengkun Wang, Binwu Wang, Yang Wang

AI总结:

TStruct是一种学习共享时间结构的网络,通过参数化共享时间依赖矩阵等方式,在多数据集上优于基线,实现长期时间序列预测的高效高精度建模。

摘要:

长期时间序列预测需要对时间依赖关系进行建模,该建模需在时间分布偏移下保持稳定且具有泛化性。然而,现实世界的时间序列通常存在噪声且非平稳,这使得深度学习模型容易拟合训练数据中的偶然时间交互。一个关键问题是,是否存在跨不同时间位置和变量重复出现的时间关系结构。我们对常用时间序列数据集上的时间关系矩阵进行分析,发现来自不同时间位置和不同变量的样本表现出共享的主导时间关系结构,这表明时间关系结构包含跨样本和变量稳定重复出现的模式。为学习此类共享结构,我们提出了时间结构网络(Temporal Structure Network,TStruct),这是一种简单却有效的网络,用于直接学习共享且可泛化的时间依赖关系。具体而言,TStruct直接参数化一组共享的时间依赖矩阵,并利用动态样本信息和静态变量先验生成归一化组合权重,这些权重表示每个样本-变量对在共享时间依赖结构上的软分配,使模型能在共享结构约束下生成自适应时间依赖矩阵,而非以完全无约束的方式拟合偶然交互。值得注意的是,这种简单的结构感知设计在所有数据集上始终优于多个强基线,在保持高计算效率的同时实现了更优的准确性。

英文摘要:

Long-term time series forecasting requires learning temporal dependencies that remain stable and generalizable under temporal distribution shifts. However, real-world time series are often noisy, non-stationary, and heterogeneous, making deep learning models prone to overfitting incidental temporal interactions in the training data. This raises a question: when observations are difficult to align, could temporal relationships be easier to reuse? To investigate this possibility, we analyze patch-level temporal relationship matrices across commonly used time series datasets, revealing dominant patterns shared across different temporal windows and variables. Comparisons with patch-shuffled and IID Gaussian controls further support the presence of reusable temporal relationship structures. Motivated by these findings, we propose \textbf{T}emporal \textbf{Struct}ure Network (\textbf{TStruct}), a simple yet effective network that learns reusable temporal dependency structures and adaptively combines them for each sample and variable. Specifically, TStruct directly parameterizes a set of shared temporal dependency matrices and generates normalized combination weights using dynamic sample information and static variable priors. These weights represent a soft allocation of each sample-variable pair over shared temporal dependency structures, enabling the model to generate adaptive temporal dependency matrices under shared structural constraints, rather than fitting incidental interactions in a fully unconstrained manner. Notably, TStruct consistently outperforms several strong baselines across all evaluated datasets, achieving superior accuracy while maintaining high computational efficiency. Our code is available at https://github.com/yangzeli/TStruct.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑