用语言提示+强化学习，让大模型更好预测时间序列。
LangTime: A Language-Guided Unified Model for Time Series Forecasting with Proximal Policy Optimization
- 设计时序理解提示，将数据压缩为单个标记，增强跨领域适应性。
- 引入多维奖励机制，使自回归预测更稳定，误差累积减少。
- 适合需要跨领域预测的工业场景，尤其关注长期稳定性。
近期研究显示，预训练大语言模型（LLMs）在时间序列应用中日益受到关注。然而，将LLMs作为时间序列预测的基础模型面临三大挑战：跨领域泛化、跨模态对齐及自回归框架中的误差累积。为此，我们提出LangTime，一种基于语言引导的统一时间序列预测模型，结合跨领域预训练与基于强化学习的微调。LangTime构建了时序理解提示（TCPs），包含数据集级和通道级指令，促进领域适应，并将时间序列压缩为单一标记，使LLM能更好地理解与对齐时序数据。为改进自回归预测，我们引入TimePPO——一种基于强化学习的微调算法。TimePPO通过针对时间序列定制的多维奖励函数和基于重复的价值估计策略，有效缓解误差累积。大量实验表明，LangTime在跨领域预测任务中达到当前最优性能，且TimePPO微调显著提升了自回归预测的稳定性和准确性。
原文摘要 · Abstract (English)
Recent research has shown an increasing interest in utilizing pre-trained large language models (LLMs) for a variety of time series applications. However, there are three main challenges when using LLMs as foundational models for time series forecasting: (1) Cross-domain generalization. (2) Cross-modality alignment. (3) Error accumulation in autoregressive frameworks. To address these challenges, we proposed LangTime, a language-guided unified model for time series forecasting that incorporates cross-domain pre-training with reinforcement learning-based fine-tuning. Specifically, LangTime constructs Temporal Comprehension Prompts (TCPs), which include dataset-wise and channel-wise instructions, to facilitate domain adaptation and condense time series into a single token, enabling LLMs to understand better and align temporal data. To improve autoregressive forecasting, we introduce TimePPO, a reinforcement learning-based fine-tuning algorithm. TimePPO mitigates error accumulation by leveraging a multidimensional rewards function tailored for time series and a repeat-based value estimation strategy. Extensive experiments demonstrate that LangTime achieves state-of-the-art cross-domain forecasting performance, while TimePPO fine-tuning effectively enhances the stability and accuracy of autoregressive forecasting.
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。