Dynamic Hypergraph Representation Learning for Multivariate Time Series without Prior Knowledge

2026-05-21

AI Summary

This work addresses the challenge of modeling high-order relationships in multivariate time series when prior structural knowledge is unavailable. To this end, the authors propose a dynamic hypergraph construction and prediction framework that does not require predefined hypergraph structures. The method automatically identifies latent groups by integrating community detection with attention mechanisms and translates these groups into dynamic hypergraphs via clique-based constructions. A novel Dynamic Hypergraph Attention Convolutional Network (DHACN) is then designed to capture temporal dynamics over such evolving hypergraphs. Experimental results across multiple real-world datasets demonstrate that the proposed approach effectively uncovers high-order dependencies and significantly improves forecasting accuracy, thereby establishing a new paradigm for learning high-order temporal relationships without prior structural assumptions.

Abstract

Hypergraphs have the capacity to capture higher-dimensional relationships among entities across various domains, making them a subject of growing interest within the research community for understanding the structure and dynamics of complex systems. However, a key challenge is the derivation of hypergraph representations from time series data in situations where the structure of the hypergraph is limited or absent. In this study, we propose a model that constructs a dynamic hypergraph representation for multivariate time series without relying on prior knowledge of the data. This is achieved by applying community detection to the time series and transforming the resulting communities, obtained through an attention mechanism, into a hypergraph using a clique-based technique. Hypergraph representations are derived from different time series datasets, and the resulting hypergraphs are then used by a Dynamic Hypergraph Attention Convolution Network (DHACN) for multivariate time series predictions. This research advances the field of hypergraph representation by introducing a novel approach that is better suited to uncover high-order relationships without prior knowledge.

Problem

• Dynamic Hypergraph
• Multivariate Time Series
• Representation Learning
• Prior Knowledge
• Higher-order Relationships

Innovation

• dynamic hypergraph
• multivariate time series
• community detection
• attention mechanism
• hypergraph representation learning