用局部结构对比提升时间序列预测准确率
Patch-wise Structural Loss for Time Series Forecasting
- 按片段比较时间序列的统计特征，捕捉点对点损失忽略的结构关系
- 在多个真实数据集上显著提升主流模型的预测性能
- 适合需要精准捕捉复杂时序模式的研究者与应用开发者
时间序列预测在机器学习中备受关注，因其在诸多领域具有关键作用。然而，现有多数预测模型依赖均方误差等点对点损失函数，将每个时间步视为独立，忽视了时间序列固有的结构依赖性，难以准确捕捉复杂的时序模式。为此，本文提出一种新型分片结构（Patch-wise Structural, PS）损失，通过在片段层面比较时间序列，利用局部统计特性（如相关性、方差、均值）来捕捉传统点对点损失忽略的细微结构差异。PS损失可与点对点损失无缝结合，同时优化局部结构不一致性和单个时间步误差。大量实验表明，该损失能显著提升多种真实数据集上先进模型的性能，为精确建模复杂时间序列数据建立了新基准，并提供了损失函数设计的新视角。
原文摘要 · Abstract (English)
Time-series forecasting has gained significant attention in machine learning due to its crucial role in various domains. However, most existing forecasting models rely heavily on point-wise loss functions like Mean Square Error, which treat each time step independently and neglect the structural dependencies inherent in time series data, making it challenging to capture complex temporal patterns accurately. To address these challenges, we propose a novel Patch-wise Structural (PS) loss, designed to enhance structural alignment by comparing time series at the patch level. Through leveraging local statistical properties, such as correlation, variance, and mean, PS loss captures nuanced structural discrepancies overlooked by traditional point-wise losses. Furthermore, it integrates seamlessly with point-wise loss, simultaneously addressing local structural inconsistencies and individual time-step errors. PS loss establishes a novel benchmark for accurately modeling complex time series data and provides a new perspective on time series loss function design. Extensive experiments demonstrate that PS loss significantly improves the performance of state-of-the-art models across diverse real-world datasets.
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。