QuantileFormer: Probabilistic Time Series Forecasting with a Pattern-Mixture Decomposed VAE Transformer

Shao, Yimiao; Li, Wenzhong; Xia, Kang; Lin, Kaijie; Lin, Mingkai; Lu, Sanglu

doi:10.24963/ijcai.2025/684

QuantileFormer: Probabilistic Time Series Forecasting with a Pattern-Mixture Decomposed VAE Transformer

Yimiao Shao, Wenzhong Li, Kang Xia, Kaijie Lin, Mingkai Lin, Sanglu Lu

Download BibTeX Cite

Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence. Main Track. Pages 6147-6155.

https://doi.org/10.24963/ijcai.2025/684

Keywords

Data Mining: DM: Mining spatial and/or temporal data
Machine Learning: ML: Time series and data streams

Abstract

Probabilistic time series forecasting has attracted an increasing attention in machine learning community for its potential applications in the fields of renewable energy, traffic management, healthcare, etc. Previous research mainly focused on extracting long-range dependencies for point-wise prediction, which fail to capture complex temporal patterns and statistical characteristics for probabilistic analysis. In this paper, we propose a novel pattern-mixture decomposition method that decomposes long-term series into quantile drift, divergence patterns, and Gaussian mixture components, which can effectively capture the intricate temporal patterns and stochastic characteristics in time series. Based on pattern-mixture decomposition, we propose a novel Transformer-based model called QuantileFormer for probabilistic time series forecasting. It takes the the comprehensive drift-divergence mixture patterns as features, and designs a variational inference based fusion Transformer architecture to generate quantile prediction results. Extensive experiments show that the proposed method consistently boosts the baseline methods by a large margin and achieves state-of-the-art performance on six real-world benchmarks.

Yimiao Shao, Wenzhong Li, Kang Xia, Kaijie Lin, Mingkai Lin, and Sanglu Lu. 2025. QuantileFormer: Probabilistic Time Series Forecasting with a Pattern-Mixture Decomposed VAE Transformer. In Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25. IJCAI, 6147–6155. https://doi.org/10.24963/ijcai.2025/684

@inproceedings{ijcai2025p684,
  title     = {QuantileFormer: Probabilistic Time Series Forecasting with a Pattern-Mixture Decomposed VAE Transformer},
  author    = {Shao, Yimiao and Li, Wenzhong and Xia, Kang and Lin, Kaijie and Lin, Mingkai and Lu, Sanglu},
  booktitle = {Proceedings of the Thirty-Fourth International Joint Conference on
               Artificial Intelligence, {IJCAI-25}},
  publisher = {International Joint Conferences on Artificial Intelligence Organization},
  editor    = {James Kwok},
  pages     = {6147--6155},
  year      = {2025},
  month     = {8},
  note      = {Main Track},
  doi       = {10.24963/ijcai.2025/684},
  url       = {https://doi.org/10.24963/ijcai.2025/684},
}

TY  - CPAPER
TI  - QuantileFormer: Probabilistic Time Series Forecasting with a Pattern-Mixture Decomposed VAE Transformer
AU  - Shao, Yimiao
AU  - Li, Wenzhong
AU  - Xia, Kang
AU  - Lin, Kaijie
AU  - Lin, Mingkai
AU  - Lu, Sanglu
T2  - Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence, IJCAI-25
PB  - International Joint Conferences on Artificial Intelligence Organization
PY  - 2025
SP  - 6147
EP  - 6155
DO  - 10.24963/ijcai.2025/684
UR  - https://doi.org/10.24963/ijcai.2025/684
ER  -