A Study of Joint Graph Inference and Forecasting
arXiv:2109.04979
Abstract
We study a recent class of models which uses graph neural networks (GNNs) to improve forecasting in multivariate time series. The core assumption behind these models is that there is a latent graph between the time series (nodes) that governs the evolution of the multivariate time series. By parameterizing a graph in a differentiable way, the models aim to improve forecasting quality. We compare four recent models of this class on the forecasting task. Further, we perform ablations to study their behavior under changing conditions, e.g., when disabling the graph-learning modules and providing the ground-truth relations instead. Based on our findings, we propose novel ways of combining the existing architectures.
arXiv commentPublished at the ICML 2021 Time Series Workshop
References in corpus (10)
- Neural Message Passing for Quantum Chemistry
- T-GCN: A Temporal Graph ConvolutionalNetwork for Traffic Prediction
- Categorical Reparameterization with Gumbel-Softmax
- The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables
- Scaling Graph Neural Networks with Approximate PageRank
- GC-LSTM: Graph Convolution Embedded LSTM for Dynamic Link Prediction
- Graph-Based Deep Modeling and Real Time Forecasting of Sparse Spatio-Temporal Data
- Forecasting: theory and practice
- A3T-GCN: Attention Temporal Graph Convolutional Network for Traffic Forecasting
- Graph Neural Network-Based Anomaly Detection in Multivariate Time Series