A Survey on Graph Neural Networks for Time Series: Forecasting, Classification, Imputation, and Anomaly Detection

Ming Jin,Huan Yee Koh,Qingsong Wen,Daniele Zambon,Cesare Alippi,Geoffrey I. Webb,Irwin King,Shirui Pan

2024-08-09

Abstract:Time series are the primary data type used to record dynamic system measurements and generated in great volume by both physical sensors and online processes (virtual sensors). Time series analytics is therefore crucial to unlocking the wealth of information implicit in available data. With the recent advancements in graph neural networks (GNNs), there has been a surge in GNN-based approaches for time series analysis. These approaches can explicitly model inter-temporal and inter-variable relationships, which traditional and other deep neural network-based methods struggle to do. In this survey, we provide a comprehensive review of graph neural networks for time series analysis (GNN4TS), encompassing four fundamental dimensions: forecasting, classification, anomaly detection, and imputation. Our aim is to guide designers and practitioners to understand, build applications, and advance research of GNN4TS. At first, we provide a comprehensive task-oriented taxonomy of GNN4TS. Then, we present and discuss representative research works and introduce mainstream applications of GNN4TS. A comprehensive discussion of potential future research directions completes the survey. This survey, for the first time, brings together a vast array of knowledge on GNN-based time series research, highlighting foundations, practical applications, and opportunities of graph neural networks for time series analysis.

Machine Learning,Artificial Intelligence

What problem does this paper attempt to address?

The main problem this paper attempts to address is the four fundamental tasks in time series analysis: forecasting, classification, anomaly detection, and imputation. Specifically: 1. **Time Series Forecasting**: Predicting future time series values based on historical observation data. Depending on the application needs, it can be divided into single-step forecasting and multi-step forecasting. 2. **Time Series Classification**: Classifying time series, such as categorizing time series into different classes or states. 3. **Anomaly Detection**: Detecting irregular and unexpected events in time series data. This includes determining when an anomaly event occurs and diagnosing the cause of the anomaly. 4. **Imputation**: Estimating and filling in missing or incomplete data points in the time series. The paper addresses these tasks through Graph Neural Networks (GNN), leveraging GNN's ability to explicitly model temporal relationships and inter-variable relationships in time series, which is challenging for traditional methods and other deep learning approaches. The paper also provides a comprehensive classification and representative research work for these tasks and discusses potential future research directions for GNN in time series analysis.

A Survey on Graph Neural Networks for Time Series: Forecasting, Classification, Imputation, and Anomaly Detection

A Systematic Literature Review of Spatio-Temporal Graph Neural Network Models for Time Series Forecasting and Classification

Graph Neural Networks for temporal graphs: State of the art, open challenges, and opportunities

A Survey on Dynamic Graph Neural Networks Modeling

A Comprehensive Survey on Graph Neural Networks

A survey of dynamic graph neural networks

Connecting the Dots: Multivariate Time Series Forecasting with Graph Neural Networks

TimeGNN: Temporal Dynamic Graph Learning for Time Series Forecasting

FourierGNN: Rethinking Multivariate Time Series Forecasting from a Pure Graph Perspective

Graph Neural Networks Designed for Different Graph Types: A Survey

Graph Neural Network for spatiotemporal data: methods and applications

Spatio-Temporal Graph Neural Networks: A Survey

Graph Anomaly Detection in Time Series: A Survey

STGSN — A Spatial–Temporal Graph Neural Network framework for time-evolving social networks

Dynamic Graph Learning-Neural Network for Multivariate Time Series Modeling

Grouped Graph Neural Networks for Anomaly Detection in Time Series

Spatio-Temporal Graph Neural Networks for Predictive Learning in Urban Computing: A Survey

Graph-enabled Reinforcement Learning for Time Series Forecasting with Adaptive Intelligence

Graph Neural Networks for Tabular Data Learning: A Survey with Taxonomy and Directions

A Comprehensive Survey of Dynamic Graph Neural Networks: Models, Frameworks, Benchmarks, Experiments and Challenges

Graph Neural Alchemist: An innovative fully modular architecture for time series-to-graph classification