MSHyper: Multi-Scale Hypergraph Transformer for Long-Range Time Series Forecasting

TL;DR

MSHyper introduces a multi-scale hypergraph transformer that effectively models high-order interactions among temporal patterns for improved long-range time series forecasting, achieving state-of-the-art results.

Contribution

The paper proposes a novel multi-scale hypergraph transformer framework that captures high-order pattern interactions for long-range forecasting, addressing limitations of previous models.

Findings

MSHyper outperforms existing methods on five real-world datasets.

The hypergraph modeling enhances the understanding of multi-scale temporal interactions.

The tri-stage message passing mechanism improves interaction learning.

Abstract

Demystifying interactions between temporal patterns of different scales is fundamental to precise long-range time series forecasting. However, previous works lack the ability to model high-order interactions. To promote more comprehensive pattern interaction modeling for long-range time series forecasting, we propose a Multi-Scale Hypergraph Transformer (MSHyper) framework. Specifically, a multi-scale hypergraph is introduced to provide foundations for modeling high-order pattern interactions. Then by treating hyperedges as nodes, we also build a hyperedge graph to enhance hypergraph modeling. In addition, a tri-stage message passing mechanism is introduced to aggregate pattern information and learn the interaction strength between temporal patterns of different scales. Extensive experiments on five real-world datasets demonstrate that MSHyper achieves state-of-the-art (SOTA) performance across various settings.