让图传播机制随时间动态变化，提升时序图学习效率与性能。
Dynamic Spectral Filtering for Temporal Graph Learning: Learning Evolving Propagation Operators
- 用时变的切比雪夫滤波器表示随时间演化的图传播算子
- 在多个数据集上达到领先或接近顶尖的预测准确率，参数量少38倍
- 适合追求计算高效、轻量化部署的时序图学习任务
时序图学习通常关注节点状态演变或交互历史编码。本文探讨一个被忽视的核心问题：图传播机制本身是否应随时间演化？提出动态谱滤波（DSF），将快照 t 时刻的传播表示为具有向量值时变系数的切比雪夫多项式滤波器。这些紧凑的多阶系数被显式建模为循环时序状态，通过循环分支更新，并由乘性全局门和阶特定门控制其幅度。时序状态与节点数量无关。在 MOOC、Wikipedia 和 Reddit 的时序链接预测基准上，收敛后的 DSF 分别取得 0.7851、0.9088、0.9860 的 AP 分数，仅需 93K 到 133K 可训练参数，峰值 GPU 内存 68 到 182 MB，每轮训练 1.6 到 2.1 秒。相比相近的 DEFT 基线，DSF 在 MOOC 上表现更优，在 Reddit 上差异小于 0.001，在 Wikipedia 上略低，但参数量减少 8.3 到 8.6 倍，GPU 内存降低 25 到 33 倍，每轮训练时间减少 5 到 19 倍。相较所有对比方法，其 GPU 内存消耗最低达 3.3 到 38.6 倍。结果表明，直接对谱响应进行演化是一种在计算效率优先时有效的时序归纳偏置。
原文摘要 · Abstract (English)
Temporal graph learning is commonly organized around the evolution of node states or the encoding of interaction histories. We study an underexplored, operator-centric question: should the graph propagation mechanism itself evolve over time? We introduce Dynamic Spectral Filtering (DSF), which represents propagation at snapshot t by a Chebyshev polynomial filter with vector-valued, time-dependent coefficients. DSF explicitly treats these compact multi-order coefficients as recurrent temporal states. A recurrent branch proposes updates, while multiplicative global and order-specific gates regulate their magnitude. The temporal state is independent of the number of nodes. On MOOC, Wikipedia, and Reddit temporal link-prediction benchmarks, converged DSF runs attain AP scores of 0.7851, 0.9088, and 0.9860, respectively, with 93K to 133K trainable parameters, 68 to 182 MB peak GPU memory, and 1.6 to 2.1 seconds of training per epoch. Against the closely related DEFT baseline, DSF is better on MOOC, within 0.001 AP on Reddit, and modestly lower on Wikipedia, while using 8.3 to 8.6 times fewer parameters, 25 to 33 times less GPU memory, and 5 to 19 times less time per epoch. Relative to all measured alternatives, it uses 3.3 to 38.6 times less GPU memory. These results support direct spectral-response evolution as a useful temporal inductive bias when computational efficiency is a first-class requirement.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。