Publication Type

Conference Proceeding Article

Version

publishedVersion

Publication Date

8-2025

Abstract

Large Language Models (LLMs) have recently shown promise in Time Series Forecasting (TSF) by effectively capturing intricate time-domain dependencies. However, our preliminary experiments reveal that standard LLM-based approaches often fail to capture global correlations, limiting predictive performance. We found that embedding frequency-domain signals smooths weight distributions and enhances structured correlations by clearly separating global trends (low-frequency components) from local variations (high-frequency components). Building on these insights, we propose FreqLLM, a novel framework that integrates frequency-domain semantic alignment into LLMs to refine prompts for improved time series analysis. By bridging the gap between frequency signals and textual embeddings, FreqLLM effectively captures multi-scale temporal patterns and provides more robust forecasting results. Extensive experiments on benchmark datasets demonstrate that FreqLLM outperforms state-of-the-art TSF methods in both accuracy and generalization. The code is available at https://github.com/biya0105/FreqLLM.

Keywords

Data mining, machine learning

Discipline

Artificial Intelligence and Robotics | Software Engineering

Research Areas

Software and Cyber-Physical Systems

Publication

Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI 2025, Montreal, August 16-22

First Page

3389

Last Page

3397

Identifier

10.24963/IJCAI.2025/377

Publisher

ijcai.org

City or Country

Montreal, Canada

Additional URL

https://doi.org/10.24963/ijcai.2025/377

Download

Share

COinS