MSGNet: Learning Multi-Scale Inter-Series Correlations for Multivariate Time Series Forecasting

被引:0
作者
Cai, Wanlin [1 ]
Liang, Yuxuan [2 ]
Liu, Xianggen [1 ]
Feng, Jianshuai [3 ]
Wu, Yuankai [1 ]
机构
[1] Sichuan Univ, Chengdu, Peoples R China
[2] Hong Kong Univ Sci & Technol Guangzhou, Guangzhou, Peoples R China
[3] Beijing Inst Technol, Beijing, Peoples R China
来源
THIRTY-EIGHTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, VOL 38 NO 10 | 2024年
基金
中国国家自然科学基金;
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Multivariate time series forecasting poses an ongoing challenge across various disciplines. Time series data often exhibit diverse intra-series and inter-series correlations, contributing to intricate and interwoven dependencies that have been the focus of numerous studies. Nevertheless, a significant research gap remains in comprehending the varying inter-series correlations across different time scales among multiple time series, an area that has received limited attention in the literature. To bridge this gap, this paper introduces MSGNet, an advanced deep learning model designed to capture the varying inter-series correlations across multiple time scales using frequency domain analysis and adaptive graph convolution. By leveraging frequency domain analysis, MS-GNet effectively extracts salient periodic patterns and decomposes the time series into distinct time scales. The model incorporates a self-attention mechanism to capture intra-series dependencies, while introducing an adaptive mixhop graph convolution layer to autonomously learn diverse interseries correlations within each time scale. Extensive experiments are conducted on several real-world datasets to showcase the effectiveness of MSGNet. Furthermore, MSGNet possesses the ability to automatically learn explainable multi-scale inter-series correlations, exhibiting strong generalization capabilities even when applied to out-of-distribution samples. Code is available at https://github.com/YoZhibo/MSGNet.
引用
收藏
页码:11141 / 11149
页数:9
相关论文
共 50 条
  • [41] Multi-scale transition matrix approach to time series
    Yuan, Qianshun
    Semba, Sherehe
    Zhang, Jing
    Weng, Tongfeng
    Gu, Changgui
    Yang, Huijie
    [J]. PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS, 2021, 578
  • [42] Multi-scale description and prediction of financial time series
    Nawroth, A. P.
    Friedrich, R.
    Peinke, J.
    [J]. NEW JOURNAL OF PHYSICS, 2010, 12
  • [43] Multi-Scale and Hidden Resolution Time Series Models
    Ferreira, Marco A. R.
    West, Mike
    Lee, Herbert K. H.
    Higdon, David M.
    [J]. BAYESIAN ANALYSIS, 2006, 1 (04): : 947 - 967
  • [44] Stationary Multi-scale Hierarchical Dilated Graph Convolution for Multivariate Time Series Anomaly Detection
    Liang, Lifang
    Qiu, Xuyi
    Zhang, Yan
    Guan, Donghai
    Zhang, Ji
    Yuan, Weiwei
    [J]. BIG DATA AND SECURITY, ICBDS 2023, PT II, 2024, 2100 : 52 - 66
  • [45] Multi Layer Kernel Learning for Time Series Forecasting
    Widodo, Agus
    Budi, Indra
    [J]. 2012 INTERNATIONAL CONFERENCE ON ADVANCED COMPUTER SCIENCE AND INFORMATION SYSTEMS, 2012, : 313 - 318
  • [46] Optimal multi-scale time series decomposition for financial forecasting using wavelet thresholding techniques
    Shin, T
    Han, I
    [J]. NEW DIRECTIONS IN ROUGH SETS, DATA MINING, AND GRANULAR-SOFT COMPUTING, 1999, 1711 : 533 - 542
  • [47] Multivariate Time Series Evapotranspiration Forecasting using Machine Learning Techniques
    Liyew, Chalachew Muluken
    Meo, Rosa
    Di Nardo, Elvira
    Ferraris, Stefano
    [J]. 38TH ANNUAL ACM SYMPOSIUM ON APPLIED COMPUTING, SAC 2023, 2023, : 377 - 380
  • [48] Memory Augmented Graph Learning Networks for Multivariate Time Series Forecasting
    Liu, Xiangyue
    Lyu, Xinqi
    Zhang, Xiangchi
    Gao, Jianliang
    Chen, Jiamin
    [J]. PROCEEDINGS OF THE 31ST ACM INTERNATIONAL CONFERENCE ON INFORMATION AND KNOWLEDGE MANAGEMENT, CIKM 2022, 2022, : 4254 - 4258
  • [49] Deep Learning for Non-stationary Multivariate Time Series Forecasting
    Almuammar, Manal
    Fasli, Maria
    [J]. 2019 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA), 2019, : 2097 - 2106
  • [50] Pseudoinverse learning of Fuzzy Cognitive Maps for multivariate time series forecasting
    Vanhoenshoven, Frank
    Napoles, Gonzalo
    Froelich, Wojciech
    Salmeron, Jose L.
    Vanhoof, Koen
    [J]. APPLIED SOFT COMPUTING, 2020, 95