Large language models are few-shot multivariate time series classifiers
Authors:
Open Access
- 7 August 2025
- journal article
- research article
- Published by Springer Nature in Data Mining and Knowledge Discovery
- Vol. 39 (5) , 1-28
Abstract
Large Language Models (LLMs) are widely applied in time series analysis. Yet, their utility in few-shot classification—a scenario with limited training data—remains unexplored. We aim to leverage the pre-trained knowledge in LLMs to overcome the data scarcity problem within multivariate time series. To this end, we propose LLMFew, an LLM-enhanced framework, to investigate the feasibility and capacity of LLMs for few-shot multivariate time series classification (MTSC). We first introduce a Patch-wise Temporal Convolution Encoder (PTCEnc) to align time series data with the textual embedding input of LLMs. Then, we fine-tune the pre-trained LLM decoder with Low-rank Adaptations (LoRA) to enable effective representation learning from time series data. Experimental results show our model consistently outperforms state-of-the-art baselines by a large margin, achieving 125.2% and 50.2% improvement in classification accuracy on Handwriting and EthanolConcentration datasets, respectively. Our results also show LLM-based methods achieve comparable performance to traditional models across various datasets in few-shot MTSC, paving the way for applying LLMs in practical scenarios where labeled data are limited. Our code is available at https://github.com/junekchen/llm-fewshot-mtsc.Keywords
All Related Versions
Funding Information
- University of Technology Sydney
This publication has 28 references indexed in Scilit:
- Dyformer: A dynamic transformer-based architecture for multivariate time series classificationInformation Sciences, 2023
- Meta-Feature Fusion for Few-Shot Time Series ClassificationIEEE Access, 2023
- Similarity Learning based Few Shot Learning for ECG Time Series Classification2021
- ROCKET: exceptionally fast and accurate time series classification using random convolutional kernelsData Mining and Knowledge Discovery, 2020
- Interpretable Time-series Classification on Few-shot Samples2020
- Decentralized Knowledge Acquisition for Mobile Internet ApplicationsWorld Wide Web, 2020
- Deep learning for time series classification: a reviewData Mining and Knowledge Discovery, 2019
- Transfer learning for time series classification2018
- LSTM Fully Convolutional Networks for Time Series ClassificationIEEE Access, 2017
- Support vector machines and dynamic time warping for time series2008