EUVIMEDEUVIMED entdecken
Crossref

TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment

Chenxi Liu, Qianxiong Xu, Hao Miao, Sun Yang, Lingzheng Zhang, Cheng Long, Ziyue Li, Rui Zhao

Proceedings of the AAAI Conference on Artificial Intelligence · 11.04.2025 · 39 (18) · 18780-18788

DOI: 10.1609/aaai.v39i18.34067 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 11.04.2025. Verfasst von Chenxi Liu und 7 weiteren Autor:innen. Die Arbeit wurde bisher 126-mal zitiert (Crossref).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Multivariate time series forecasting (MTSF) aims to learn temporal dynamics among variables to forecast future time series. Existing statistical and deep learning-based methods suffer from limited learnable parameters and small-scale training data. Recently, large language models (LLMs) combining time series with textual prompts have achieved promising performance in MTSF. However, we discovered that current LLM-based solutions fall short in learning disentangled embeddings. We introduce TimeCMA, an intuitive yet effective framework for MTSF via cross-modality alignment. Specifically, we present a dual-modality encoding with two branches: the time series encoding branch extracts disentangled yet weak time series embeddings, and the LLM-empowered encoding branch wraps the same time series with text as prompts to obtain entangled yet robust prompt embeddings. As a result, such a cross-modality alignment retrieves both disentangled and robust time series embeddings, ``the best of two worlds'', from the prompt embeddings based on time series and prompt modality similarities. As another key design, to reduce the computational costs from time series with their length textual prompts, we design an effective prompt to encourage the most essential temporal information to be encapsulated in the last token: only the last token is passed to downstream prediction. We further store the last token embeddings to accelerate inference speed. Extensive experiments on eight real datasets demonstrate that TimeCMA outperforms state-of-the-arts.

Zitieren

APA

Liu, C., Xu, Q., Miao, H., Yang, S., Zhang, L., Long, C., Li, Z., & Zhao, R. (2025). TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment. Proceedings of the AAAI Conference on Artificial Intelligence, 39(18), 18780-18788. https://doi.org/10.1609/aaai.v39i18.34067

Vancouver

Liu C, Xu Q, Miao H, Yang S, Zhang L, Long C, et al. TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment. Proceedings of the AAAI Conference on Artificial Intelligence. 2025;39(18):18780-18788. doi:10.1609/aaai.v39i18.34067

BibTeX
@article{Liu2025,
  author = {Liu, Chenxi and Xu, Qianxiong and Miao, Hao and Yang, Sun and Zhang, Lingzheng and Long, Cheng and Li, Ziyue and Zhao, Rui},
  title = {TimeCMA: Towards LLM-Empowered Multivariate Time Series Forecasting via Cross-Modality Alignment},
  journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
  year = {2025},
  volume = {39},
  number = {18},
  pages = {18780--18788},
  doi = {10.1609/aaai.v39i18.34067}
}
Kostenlos starten →