GCformer: An Efficient Framework for Accurate and Scalable Long-Term Multivariate Time Series Forecasting

被引:9
|
作者
Zhao, Yanjun [1 ,2 ]
Ma, Ziqing [2 ]
Zhou, Tian [2 ]
Ye, Mengni [2 ]
Sun, Liang [2 ]
Qian, Yi [1 ]
机构
[1] Xi An Jiao Tong Univ, Xian, Peoples R China
[2] Alibaba Grp, Hangzhou, Peoples R China
来源
PROCEEDINGS OF THE 32ND ACM INTERNATIONAL CONFERENCE ON INFORMATION AND KNOWLEDGE MANAGEMENT, CIKM 2023 | 2023年
关键词
Global Convolution Kernel; Transformer; Global-Local Design; Time Series Forecasting;
D O I
10.1145/3583780.3615136
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Transformer-based models have emerged as promising tools for time series forecasting. However, these models cannot make accurate prediction for long input time series. On the one hand, they failed to capture long-range dependency within time series data. On the other hand, the long input sequence usually leads to large model size and high time complexity. To address these limitations, we present GCformer, which combines a structured global convolutional branch for processing long input sequences with a local Transformer-based branch for capturing short, recent signals. A cohesive framework for a global convolution kernel has been introduced, utilizing three distinct parameterization methods. The selected structured convolutional kernel in the global branch has been specifically crafted with sublinear complexity, thereby allowing for the efficient and effective processing of lengthy and noisy input signals. Empirical studies on six benchmark datasets demonstrate that GCformer outperforms state-of-the-art methods, reducing MSE error in multivariate time series benchmarks by 4.38% and model parameters by 61.92%. In particular, the global convolutional branch can serve as a plug-in block to enhance the performance of other models, with an average improvement of 31.93%, including various recently published Transformer-based models. Our code is publicly available at https://github.com/Yanjun-Zhao/GCformer.
引用
收藏
页码:3464 / 3473
页数:10
相关论文
共 50 条
  • [1] SageFormer: Series-Aware Framework for Long-Term Multivariate Time-Series Forecasting
    Zhang, Zhenwei
    Meng, Linghang
    Gu, Yuantao
    IEEE INTERNET OF THINGS JOURNAL, 2024, 11 (10): : 18435 - 18448
  • [2] Hierarchical attention network for multivariate time series long-term forecasting
    Bi, Hongjing
    Lu, Lilei
    Meng, Yizhen
    APPLIED INTELLIGENCE, 2023, 53 (05) : 5060 - 5071
  • [3] Hierarchical attention network for multivariate time series long-term forecasting
    Hongjing Bi
    Lilei Lu
    Yizhen Meng
    Applied Intelligence, 2023, 53 : 5060 - 5071
  • [4] VAECGAN: a generating framework for long-term prediction in multivariate time series
    Xiang Yin
    Yanni Han
    Zhen Xu
    Jie Liu
    Cybersecurity, 4
  • [5] VAECGAN: a generating framework for long-term prediction in multivariate time series
    Yin, Xiang
    Han, Yanni
    Xu, Zhen
    Liu, Jie
    CYBERSECURITY, 2021, 4 (01)
  • [6] CNformer: a convolutional transformer with decomposition for long-term multivariate time series forecasting
    Xingyu Wang
    Hui Liu
    Zhihan Yang
    Junzhao Du
    Xiyao Dong
    Applied Intelligence, 2023, 53 : 20191 - 20205
  • [7] CNformer: a convolutional transformer with decomposition for long-term multivariate time series forecasting
    Wang, Xingyu
    Liu, Hui
    Yang, Zhihan
    Du, Junzhao
    Dong, Xiyao
    APPLIED INTELLIGENCE, 2023, 53 (17) : 20191 - 20205
  • [8] Enhancing Long-Term Cloud Workload Forecasting Framework: Anomaly Handling and Ensemble Learning in Multivariate Time Series
    Kim, Yeong-Min
    Song, Seunghwan
    Koo, Byoung-Mo
    Son, Jeena
    Lee, Yeseul
    Baek, Jun-Geol
    IEEE TRANSACTIONS ON CLOUD COMPUTING, 2024, 12 (02) : 789 - 799
  • [9] Representing Multiview Time-Series Graph Structures for Multivariate Long-Term Time-Series Forecasting
    Wang Z.
    Fan J.
    Wu H.
    Sun D.
    Wu J.
    IEEE Transactions on Artificial Intelligence, 5 (06): : 2651 - 2662
  • [10] MSDformer: an autocorrelation transformer with multiscale decomposition for long-term multivariate time series forecasting
    Su, Guangyao
    Guan, Yepeng
    APPLIED INTELLIGENCE, 2025, 55 (02)