用通道描述训练时间序列基础模型，实现跨任务迁移
Time to Embed: Unlocking Foundation Models for Time Series with Channel Descriptions
- 引入通道文本描述，构建对通道顺序不变的时序嵌入模型
- 700万参数模型在多任务上达顶尖性能，超越现有方法
- 适合需要低依赖特征工程的时序分析场景
传统时间序列模型通常任务专用，依赖数据集特有训练和大量特征工程。尽管基于Transformer的架构提升了可扩展性，但如文本、视觉和音频中常见的基础模型在时序领域仍研究不足，且主要局限于预测任务。我们提出CHARM，一种用于多元时间序列的基础嵌入模型，可学习共享、可迁移且具备领域感知的表示。为应对时序基础学习的独特挑战，CHARM在架构上创新，融合通道级文本描述，同时保持对通道顺序的不变性。模型采用联合嵌入预测架构（JEPA）进行训练，结合新颖的数据增强策略与损失函数，提升可解释性与训练稳定性。我们的700万参数模型在多种下游任务中表现领先，为时序表征学习树立了新基准。
原文摘要 · Abstract (English)
Traditional time series models are task-specific and often depend on dataset-specific training and extensive feature engineering. While Transformer-based architectures have improved scalability, foundation models, commonplace in text, vision, and audio, remain under-explored for time series and are largely restricted to forecasting. We introduce $\textbf{CHARM}$, a foundation embedding model for multivariate time series that learns shared, transferable, and domain-aware representations. To address the unique difficulties of time series foundation learning, $\textbf{CHARM}$ incorporates architectural innovations that integrate channel-level textual descriptions while remaining invariant to channel order. The model is trained using a Joint Embedding Predictive Architecture (JEPA), with novel augmentation schemes and a loss function designed to improve interpretability and training stability. Our $7$M-parameter model achieves state-of-the-art performance across diverse downstream tasks, setting a new benchmark for time series representation learning.
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。