Graph Neural Networks for Multivariate Time-Series Forecasting via Learning Hierarchical Spatiotemporal Dependencies
Primary Area: representation learning for computer vision, audio, language, and other modalities
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Keywords: Multivariate time-series forecasting, Spatiotemporal graph neural networks, Deep learning
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
TL;DR: We propose HSDGNN, a hierarchical spatiotemporal dependencies learning based graph neural network model that leverages spatial-, temporal-, and intra-dependency learning in a unified framework.
Abstract: Multivariate time-series forecasting is one of the essential tasks to draw insights from sequential data. Spatiotemporal Graph Neural Networks (STGNNs) have attracted much attention in this field due to their capability to capture the underlying spatiotemporal dependencies. However, current STGNN solutions still fall short of providing trustworthy predictions due to insufficient modeling of the dependencies and dynamics at different levels. In this paper, we propose a graph neural network model for multivariate time-series forecasting via learning hierarchical spatiotemporal dependencies (HSDGNN). Specifically, we organize variables as nodes in a graph while each node serves as a subgraph consisting of the attributes of variables. Then we design two-level convolutions on the hierarchical graph to model the spatial dependencies with different granularities. The changes in graph topologies are also encoded for strengthening dependency modeling across time and spatial dimensions. We test the proposed model on real-world datasets from different domains. The experimental results demonstrate the superiority of HSDGNN over state-of-the-art baselines in terms of prediction accuracy.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 3207
Loading