A Comprehensive Survey of Deep Learning for Multivariate Time Series Forecasting: A Channel Strategy Perspective

2025-02-15

AI Summary

In multivariate time series forecasting (MTSF), insufficient modeling of inter-channel dependencies limits prediction accuracy. This paper presents a systematic survey of deep learning–based channel modeling approaches and proposes the first three-dimensional taxonomy—structured along strategy, mechanism, and property—grounded in the intrinsic nature of information interaction. The taxonomy unifies diverse paradigms, including coupled/decoupled architectures, graph neural networks, attention mechanisms, meta-learning, and interpretability methods. We construct an extensible evaluation framework covering end-to-end modeling, feature enhancement, and dynamic weight allocation pathways; synthesize over 100 works; identify critical bottlenecks; and publicly release a continuously updated open-source knowledge base on GitHub. Our work provides theoretical guidance for algorithm design and establishes practical benchmarks for real-world deployment.

Technology Category

Machine Learning: Deep Neural Architectures and Foundation ModelsPlanning, Routing, and Scheduling: Optimization of Spatio-temporal SystemsMultiagent Systems: Other Foundations of Multi Agent Systems

Application Category

Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphsSemantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semanticsWeb Mining and Content Analysis: Models for Web evolution

Abstract

Multivariate Time Series Forecasting (MTSF) plays a crucial role across diverse fields, ranging from economic, energy, to traffic. In recent years, deep learning has demonstrated outstanding performance in MTSF tasks. In MTSF, modeling the correlations among different channels is critical, as leveraging information from other related channels can significantly improve the prediction accuracy of a specific channel. This study systematically reviews the channel modeling strategies for time series and proposes a taxonomy organized into three hierarchical levels: the strategy perspective, the mechanism perspective, and the characteristic perspective. On this basis, we provide a structured analysis of these methods and conduct an in-depth examination of the advantages and limitations of different channel strategies. Finally, we summarize and discuss some future research directions to provide useful research guidance. Moreover, we maintain an up-to-date Github repository (https://github.com/decisionintelligence/CS4TS) which includes all the papers discussed in the survey.

Problem

• Explores deep learning for multivariate time series forecasting.
• Analyzes channel correlation strategies in time series modeling.
• Proposes taxonomy and reviews future research directions.

Innovation

• Deep learning for time series
• Channel correlation modeling
• Structured taxonomy analysis