针对时间序列模型微调，提出多尺度方法提升性能。
Multi-Scale Finetuning for Encoder-based Time Series Foundation Models
- 在微调中显式引入多尺度建模机制。
- 在三种骨干模型上超越现有方法，最高提升12.3%。
- 适合需要高精度预测的时间序列任务研究者。
时间序列基础模型（TSFMs）在零样本时间序列预测中表现出色。然而，如何有效在下游任务上微调这些模型仍是一个重要但未被充分探索的挑战。简单微调虽能带来性能提升，但难以充分发挥TSFMs的能力，常导致过拟合和次优表现。鉴于不同采样尺度下的多样时序模式及TSFMs固有的多尺度预测能力，我们从因果视角分析微调过程，揭示了显式建模多尺度的重要性，并指出传统方法的不足。针对编码器型TSFMs，我们提出多尺度微调（MSFT），一种简单且通用的框架，将多尺度建模明确融入微调流程。在Moirai、Moment和Units三个不同骨干模型上的实验表明，采用MSFT微调的TSFMs不仅优于简单微调和典型的参数高效微调方法，还超越了当前最优深度学习方法。代码已开源。
原文摘要 · Abstract (English)
Time series foundation models (TSFMs) demonstrate impressive zero-shot performance for time series forecasting. However, an important yet underexplored challenge is how to effectively finetune TSFMs on specific downstream tasks. While naive finetuning can yield performance gains, we argue that it falls short of fully leveraging TSFMs' capabilities, often resulting in overfitting and suboptimal performance. Given the diverse temporal patterns across sampling scales and the inherent multi-scale forecasting capabilities of TSFMs, we adopt a causal perspective to analyze finetuning process, through which we highlight the critical importance of explicitly modeling multiple scales and reveal the shortcomings of naive approaches. Focusing on encoder-based TSFMs, we propose Multiscale finetuning (MSFT), a simple yet general framework that explicitly integrates multi-scale modeling into the finetuning process. Experimental results on three different backbones (Moirai, Moment and Units) demonstrate that TSFMs finetuned with MSFT not only outperform naive and typical parameter efficient finetuning methods but also surpass state-of-the-art deep learning methods. Codes are available at https://github.com/zqiao11/MSFT.
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。