Multivariate Time Series Forecasting With Dynamic Graph Neural ODEs

Multivariate time series forecasting has long received significant attention in real-world applications, such as energy consumption and traffic prediction. While recent methods demonstrate good forecasting abilities, they have three fundamental limitations. (i). Discrete neural architectures: Interl...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on knowledge and data engineering Vol. 35; no. 9; pp. 9168 - 9180
Main Authors: Jin, Ming, Zheng, Yu, Li, Yuan-Fang, Chen, Siheng, Yang, Bin, Pan, Shirui
Format: Journal Article
Language:English
Published: New York IEEE 01.09.2023
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Computational modeling
Electronic mail
Energy consumption
Forecasting
graph neural networks
Mathematical models
Message passing
Multivariate analysis
Multivariate time series forecasting
neural ordinary differential equations
Predictive models
Time series
Time series analysis
Topology
Trajectory
ISSN:1041-4347, 1558-2191
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Multivariate time series forecasting has long received significant attention in real-world applications, such as energy consumption and traffic prediction. While recent methods demonstrate good forecasting abilities, they have three fundamental limitations. (i). Discrete neural architectures: Interlacing individually parameterized spatial and temporal blocks to encode rich underlying patterns leads to discontinuous latent state trajectories and higher forecasting numerical errors. (ii). High complexity: Discrete approaches complicate models with dedicated designs and redundant parameters, leading to higher computational and memory overheads. (iii). Reliance on graph priors: Relying on predefined static graph structures limits their effectiveness and practicability in real-world applications. In this paper, we address all the above limitations by proposing a continuous model to forecast M ultivariate T ime series with dynamic G raph neural O rdinary D ifferential E quations ( MTGODE ). Specifically, we first abstract multivariate time series into dynamic graphs with time-evolving node features and unknown graph structures. Then, we design and solve a neural ODE to complement missing graph topologies and unify both spatial and temporal message passing, allowing deeper graph propagation and fine-grained temporal information aggregation to characterize stable and precise latent spatial-temporal dynamics. Our experiments demonstrate the superiorities of MTGODE from various perspectives on five time series benchmark datasets.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1041-4347
1558-2191
DOI:10.1109/TKDE.2022.3221989