面向不规则多元时间序列预测的规范化预对齐方法复兴
Revitalizing Canonical Pre-Alignment for Irregular Multivariate Time Series Forecasting
AI总结：
针对不规则多元时间序列预测中CPA预对齐计算开销大、现有图模型难以捕捉全局相关性的问题，提出基于CPA的KAFNet架构，以更低参数量和更快速度取得SOTA性能。
摘要：
以采样不均匀、变量间异步为特征的不规则多元时间序列（IMTS）支撑着大量预测应用，但对其进行高效建模仍存在挑战。规范化预对齐（CPA）通过在每个全局时间戳补零的方式被广泛应用于IMTS建模，可缓解变量间异步问题并统一序列长度，但其密集补零操作会拉长预对齐后的序列长度，尤其在变量数量较多时会带来难以承受的计算开销。近期采用分块策略的基于图的模型绕过了CPA，但这类模型的局部消息传递机制难以捕捉全局的变量间相关性。因此我们认为应当保留CPA，由模型对预对齐序列进行恰当处理，使其性能优于绕过CPA的当前最优图基线模型。技术层面，我们提出了基于CPA的紧凑IMTS预测架构KAFNet，该架构包含三个核心模块：一是用于序列平滑与缓解稀疏性的预卷积模块，二是用于序列内不规则性可学习压缩与建模的时间核聚合模块，三是用于在频域中低成本建模序列间相关性的频率线性注意力块。在多个IMTS数据集上的实验表明，KAFNet取得了当前最优的预测性能，同时参数量降低7.2倍，训练推理速度提升8.4倍。
英文摘要：
Irregular multivariate time series (IMTS), characterized by uneven sampling and inter-variate asynchrony, fuel many forecasting applications yet remain challenging to model efficiently. Canonical Pre-Alignment (CPA) has been widely adopted in IMTS modeling by padding zeros at every global timestamp, thereby alleviating inter-variate asynchrony and unifying the series length, but its dense zero-padding inflates the pre-aligned series length, especially when numerous variates are present, causing prohibitive compute overhead. Recent graph-based models with patching strategies sidestep CPA, but their local message passing struggles to capture global inter-variate correlations. Therefore, we posit that CPA should be retained, with the pre-aligned series properly handled by the model, enabling it to outperform state-of-the-art graph-based baselines that sidestep CPA. Technically, we propose KAFNet, a compact architecture grounded in CPA for IMTS forecasting that couples (1) Pre-Convolution module for sequence smoothing and sparsity mitigation, (2) Temporal Kernel Aggregation module for learnable compression and modeling of intra-series irregularity, and (3) Frequency Linear Attention blocks for the low-cost inter-series correlations modeling in the frequency domain. Experiments on multiple IMTS datasets show that KAFNet achieves state-of-the-art forecasting performance, with a 7.2$\times$ parameter reduction and a 8.4$\times$ training-inference acceleration.
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。
