发表机构
Karlsruhe Institute of Technology; The Hong Kong University of Science and Technology; East China Normal University; University of Michigan（卡尔斯鲁厄理工学院; 香港科技大学; 华东师范大学; 密歇根大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出时间相关性波动性（TCV）指标，发现高TCV下GNN等模型泛化差，进而提出GLIDE层，在合成与真实基准上使平均性能最高提升45.6%。
AI 中文摘要
将多元时间序列表示为图（其中单个序列作为节点，成对时间相关性作为边）来建模的方法已获得广泛关注。图神经网络（GNN）的近期进展通过假设图拓扑为静态并聚合相邻序列的信息，展现出了出色的性能。在本研究中，我们调查了GNN在静态和动态设置（即成对相关性随时间发生剧烈演变）下用于预测的表示能力，并确定了当前架构的关键局限。为将此形式化，我们首先提出时间相关性波动性（TCV），这是一种与模型无关的指标，旨在量化这些潜在结构的分布演变。我们建立了TCV与性能下降之间的明确关联，证明包括Transformer在内的许多流行模型在高TCV设置下泛化能力差，且常常被简单的与结构无关的基线模型超越。为解决这些局限，我们提出动态环境推理图层（GLIDE），这是一种新型GNN层，由两种基于理论的设计机制增强：（D1）基于路径的消息传递，用于捕获基于路径的邻域；（D2）静态与动态传播分离，通过局部静态近似识别最优动态性。这些组件在动态拓扑下显著提升了学习效果，同时在静态场景中保持了鲁棒性。在合成数据集和真实世界基准上的大量实验表明，GLIDE在静态和动态设置下的平均性能提升最高达45.6%，最大增益达到85.7%。源代码可在此URL获取。
英文摘要
Modeling multivariate time series by representing them as graphs, where individual series act as nodes and pairwise temporal corre- lations serve as edges, has gained significant traction. Recent advances in Graph Neural Networks (GNNs) have demonstrated strong perfor- mance by assuming a static graph topology and aggregating information from neighboring series. In this work, we investigate the representa- tional power of GNNs for forecasting under both static and dynamic settings (i.e., when pairwise correlations evolve drastically over time) and identify critical limitations in current architectures. To formalize this, we first propose Temporal Correlation Volatility (TCV), a model- agnostic metric designed to quantify the distributional evolution of these latent structures. We establish a clear connection between TCV and performance degradation, demonstrating that many popular models, including Transformers, generalize poorly in high-TCV settings and are often outperformed by simple structure-agnostic baselines. To address these limitations, we propose Graph Layer for Inference in Dynamic En- vironments (GLIDE), a novel GNN layer enhanced by two theoretically grounded design mechanisms: (D1) Path-based Message Passing, which captures path-based neighborhoods and (D2) Static and Dynamic Propagation Separation, which identifies optimal dynamics via local static approximation. These components significantly improve learning under dynamic topology while preserving robustness in static scenarios. Ex- tensive experiments on synthetic and real-world benchmarks show that GLIDE improves average performance by up to 45.6% across static and dynamic settings, with the largest gain reaching 85.7%. The source code is available at https://github.com/ChenS676/GLIDE.
Comments6 figures, 3 tables, 16 pages
Journal refEuropean Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2026
