SPHINX: Structural Prediction using Hypergraph Inference Network

TL;DR

SPHINX is a novel model that infers latent hypergraph structures from node signals, enabling better higher-order interaction modeling in domains where explicit hypergraph annotations are unavailable.

Contribution

The paper introduces an unsupervised hypergraph inference method using differentiable clustering and $k$-subset sampling, improving hypergraph structure learning for neural networks.

Findings

Successfully infers interpretable hypergraphs from data.

Enhances hypergraph neural network performance on trajectory prediction.

Addresses training stability issues with $k$-subset sampling.

Abstract

The importance of higher-order relations is widely recognized in a large number of real-world systems. However, annotating them is a tedious and sometimes impossible task. Consequently, current approaches for data modelling either ignore the higher-order interactions altogether or simplify them into pairwise connections. In order to facilitate higher-order processing, even when a hypergraph structure is not available, we introduce Structural Prediction using Hypergraph Inference Network (SPHINX), a model that learns to infer a latent hypergraph structure in an unsupervised way, solely from the final node-level signal. The model consists of a soft, differentiable clustering method used to sequentially predict, for each hyperedge, the probability distribution over the nodes and a sampling algorithm that converts them into an explicit hypergraph structure. We show that the recent advancement in $k$-subset sampling represents a suitable tool for producing discrete hypergraph structures, addressing some of the training instabilities exhibited by prior works. The resulting model can generate the higher-order structure necessary for any modern hypergraph neural network, facilitating the capture of higher-order interaction in domains where annotating them is difficult. Through extensive ablation studies and experiments conducted on two challenging datasets for trajectory prediction, we demonstrate that our model is capable of inferring suitable latent hypergraphs, that are interpretable and enhance the final performance.