Hypergraph Neural Networks
By Yifan Feng, Haoxuan You, Zizhao Zhang, Rongrong Ji, Yue Gao. AAAI Conference on Artificial Intelligence, 2018
2,070 citations recorded by the source corpus.
In this paper, we present a hypergraph neural networks (HGNN) framework for data representation learning, which can encode high-order data correlation in a hypergraph structure. Confronting the challenges of learning representation for complex data in real practice, we propose to incorporate such data structure in a hypergraph, which is more flexible on data modeling, especially when dealing with complex data. In this method, a hyperedge convolution operation is designed to handle the data correlation during representation learning. In this way, traditional hypergraph learning procedure can be conducted using hyperedge convolution operations efficiently. HGNN is able to learn the hidden layer representation considering the high-order data structure, which is a general framework considering the complex data correlations. We have conducted experiments on citation network classification and visual object recognition tasks and compared HGNN with graph convolutional networks and other traditional methods. Experimental results demonstrate that the proposed HGNN method outperforms recent state-of-theart methods. We can also reveal from the results that the proposed HGNN is superior when dealing with multi-modal data compared with existing methods.
Subfields: Artificial Intelligence, Neurology, Urban Studies
View Paper: https://www.semanticscholar.org/paper/510d98681e5e85fb1265513728f16e2543ae1b4b
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
In Neurology
- Philip H. S. Torr
- Dacheng Tao
- Yu Qiao
- Jyoti Prakash Sahoo
- Peng Gao
- Rongrong Ji
- Jia Deng
- Yisheng Song
In Urban Studies
Citations in the corpus
- HyperGCN: A New Method of Training Graph Convolutional Networks on Hypergraphs (2018, methodology): “While the state-of-the-art hypergraph neural networks (HGNN) [17] approximates each hyperedge by a clique and hence requires s C 2 (quadratic number of) edges for each hyperedge of size s , our method, i.e. HyperGCN, requires a linear number of edges (i.e.”
- A Comprehensive Survey of Few-shot Learning: Evolution, Applications, Challenges, and Opportunities (2022, methodology): “On this basis, by designing diferent embedding nodes [194ś196] and label propagation mechanisms [197, 198], graph neural networks can model the relationships between supported query pairs quite competently.”
- Hypergraph Convolution and Hypergraph Attention (2019, methodology): “Following (Defferrard et al., 2016), (Feng et al., 2019) propose a hypergraph neural network using a Cheby-shev expansion of the graph Laplacian.”
- Self-Supervised Multi-Channel Hypergraph Convolutional Network for Social Recommendation (2021, background): “HGNN [10] is the first work that designs a hyperedge convolution operation to handle complex data correlation in representation learning from a spectral perspective.”
- A Survey on Graph Neural Networks for Time Series: Forecasting, Classification, Imputation, and Anomaly Detection (2023, background): “Additionally, recent approaches based on graph transformer [100] or hypergraphs [121], such as [122] and [110], can capture longer-range spatial dependencies due to their global receptive field, making them a separate branch of enhanced methods.”
- Weisfeiler and Lehman Go Topological: Message Passing Simplicial Networks (2021, background): “We conclude this part by mentioning a large body of work on neural networks for hypergraphs, which subsume simplicial complexes (Yadati et al., 2019; Feng et al., 2019; Zhang et al., 2020; Yadati, 2020).”
- Next-item Recommendation with Sequential Hypergraphs (2020, methodology): “To capture these changes in item semantics, we propose to model such short-term item correlations in a hypergraph [1, 11], in which each hyperedge can connect multiple nodes on a single edge.”
- PointCLIP: Point Cloud Understanding by CLIP (2021, methodology): “Therein, multiview methods project point clouds onto images of multiple views and process them with 2D Convolution Neural Networks (CNN) [22] pre-trained on ImageNet [28], such as MVCNN [49] and others [14, 15, 21, 26, 62].”
- PF-Net: Point Fractal Network for 3D Point Cloud Completion (2020, background): “However, multi-view and voxel-based representation[18, 7, 34, 11, 13, 37] leads to unnecessarily voluminous and limits the output resolution of voxels.”
- Revisiting Point Cloud Shape Classification with a Simple and Effective Baseline (2021, background): “, 2018), and hypergraph methods that consider the correlation across training samples (Zhang et al., 2018; Feng et al., 2019).”
- HGNN+: General Hypergraph Neural Networks (2022, methodology): “For example, in our conference version Feng et al. [6] proposed hypergraph neural networks (HGNN) 1. https://github.com/iMoonLab/THU-DeepHypergraph towards modeling and learning beyond-pairwise complex correlations.”
- A Gentle Introduction to Deep Learning for Graphs (2019, background): “Some works on learning from hypergraph have recently been published [146, 145, 38, 67], and the most recent ones take inspiration from local and iterative processing of graphs.”
- Distilling Knowledge From Graph Convolutional Networks (2020, methodology): “Moreover, GCN based methods [8, 19, 37, 18, 41], which can directly encode the structure information from the set of points, become one of the most popular directions along this line.”
- Deep Learning for 3D Point Clouds: A Survey (2019, methodology): “Feng et al. [74] proposed a Hypergraph Neural Network (HGNN) and built a hyperedge convolutional layer by applying spectral convolution on the hypergraph.”
- Relational Representation Learning for Dynamic (Knowledge) Graphs: A Survey (2019, background): “Some recent works study ways of handling such hyperedges for static (knowledge) graphs [233, 73, 239, 10].”
- View-GCN: View-Based Graph Convolutional Network for 3D Shape Analysis (2020, neutral): “By increasing sampling rate s to 0.6 and 0.7, the view-GCNs are deeper with 4 and 5 levels having [20,12,7,4], [20,14,10,7,5] nodes respectively, and accuracies slightly drop by (0.1%,0.2”
- YOLOv13: Real-Time Object Detection with Hypergraph-Enhanced Adaptive Visual Perception (2025, neutral): “[39] proposed spectral-domain HGNNs, demonstrating the advantage in the visual retrieval task.”