Skip to main navigation Skip to search Skip to main content

LoFT-LLM: Low-Frequency Time-series Forecasting with Large Language Models

  • Jiacheng You
  • , Jingcheng Yang
  • , Yuhang Xie
  • , Zhongxuan Wu
  • , Xiucheng Li
  • , Feng Li
  • , Pengjie Wang
  • , Jian Xu
  • , Bo Zheng
  • , Xinyang Chen*
  • *Corresponding author for this work
  • Harbin Institute of Technology
  • Unaffiliated
  • Chinese University of Hong Kong

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Time-series forecasting in real-world applications such as finance and energy often faces challenges due to limited training data and complex, noisy temporal dynamics. Existing deep forecasting models typically supervise predictions using full-length temporal windows, which include substantial high-frequency noise and obscure long-term trends. Moreover, auxiliary variables containing rich domain-specific information are often underutilized, especially in few-shot settings. To address these challenges, we propose LoFT-LLM, a frequency-aware forecasting pipeline that integrates low-frequency learning with semantic calibration via a large language model (LLM). Firstly, a Patch Low-Frequency forecasting Module (PLFM) extracts stable low-frequency trends from localized spectral patches. Secondly, a residual learner then models high-frequency variations. Finally, a fine-tuned LLM refines the predictions by incorporating auxiliary context and domain knowledge through structured natural language prompts. Extensive experiments on financial and energy datasets demonstrate that LoFT-LLM significantly outperforms strong baselines under both full-data and few-shot regimes, delivering superior accuracy, robustness, and interpretability.

Original languageEnglish
Title of host publicationKDD 2026 - Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1
PublisherAssociation for Computing Machinery
Pages1809-1820
Number of pages12
ISBN (Electronic)9798400722585
DOIs
StatePublished - 20 Apr 2026
Externally publishedYes
Event32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1, KDD 2026 - Jeju Island, Korea, Republic of
Duration: 9 Aug 202613 Aug 2026

Publication series

NameProceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
Volume1-A
ISSN (Print)2154-817X

Conference

Conference32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1, KDD 2026
Country/TerritoryKorea, Republic of
CityJeju Island
Period9/08/2613/08/26

Keywords

  • few shot
  • large language models
  • low-frequency learning
  • time series forecasting

Fingerprint

Dive into the research topics of 'LoFT-LLM: Low-Frequency Time-series Forecasting with Large Language Models'. Together they form a unique fingerprint.

Cite this