Homophily-Related: Adaptive Hybrid Graph Filter for Multi-View Graph Clustering (2401.02682v1)
Abstract: Recently there is a growing focus on graph data, and multi-view graph clustering has become a popular area of research interest. Most of the existing methods are only applicable to homophilous graphs, yet the extensive real-world graph data can hardly fulfill the homophily assumption, where the connected nodes tend to belong to the same class. Several studies have pointed out that the poor performance on heterophilous graphs is actually due to the fact that conventional graph neural networks (GNNs), which are essentially low-pass filters, discard information other than the low-frequency information on the graph. Nevertheless, on certain graphs, particularly heterophilous ones, neglecting high-frequency information and focusing solely on low-frequency information impedes the learning of node representations. To break this limitation, our motivation is to perform graph filtering that is closely related to the homophily degree of the given graph, with the aim of fully leveraging both low-frequency and high-frequency signals to learn distinguishable node embedding. In this work, we propose Adaptive Hybrid Graph Filter for Multi-View Graph Clustering (AHGFC). Specifically, a graph joint process and graph joint aggregation matrix are first designed by using the intrinsic node features and adjacency relationship, which makes the low and high-frequency signals on the graph more distinguishable. Then we design an adaptive hybrid graph filter that is related to the homophily degree, which learns the node embedding based on the graph joint aggregation matrix. After that, the node embedding of each view is weighted and fused into a consensus embedding for the downstream task. Experimental results show that our proposed model performs well on six datasets containing homophilous and heterophilous graphs.
- Pseudo-labeling and confirmation bias in deep semi-supervised learning. In IJCNN, 1–8.
- Beyond low-frequency information in graph convolutional networks. In AAAI, 3950–3957.
- Simplified graph convolution with heterophily. In NeurIPS, 27184–27197.
- Variational Graph Generator for Multi-View Graph Clustering. arXiv preprint arXiv:2210.07011.
- Chung, F. R. 1997. Spectral graph theory, volume 92. American Mathematical Soc.
- A Novel Approach for Effective Multi-View Clustering with Information-Theoretic Perspective. In NeurIPS.
- Deep Multi-view Subspace Clustering with Anchor Graph. In IJCAI, 3577–3585.
- Local smoothness of graph signals. Mathematical Problems in Engineering, 2019.
- One2Multi Graph Autoencoder for Multi-view Graph Clustering. In WWW, 3070–3076.
- Hamilton, W. L. 2020. Graph Representation Learning. Synthesis Lectures on Artificial Intelligence and Machine Learning, 14(3): 1–159.
- Contrastive multi-view representation learning on graphs. In ICML, 4116–4126.
- Co-embedding: a semi-supervised multi-view representation learning approach. Neural Computing and Applications, 34: 4437 – 4457.
- Semi-supervised multi-view deep discriminant representation learning. TPAMI, 43(7): 2496–2509.
- Variational graph auto-encoders. arXiv preprint arXiv:1611.07308.
- Label efficient semi-supervised learning via graph filtering. In CVPR, 9582–9591.
- Beyond low-pass filters: Adaptive feature propagation on graphs. In Machine Learning and Knowledge Discovery in Databases, 450–465.
- Graph Filter-based Multi-view Attributed Graph Clustering. In IJCAI, 2723–2729.
- Dual Label-Guided Graph Refinement for Multi-View Graph Clustering. In AAAI, 8791–8798.
- Beyond Smoothing: Unsupervised Graph Representation Learning with Edge Heterophily Discriminating. In AAAI, 4516–4524.
- Revisiting Heterophily For Graph Neural Networks. In NeurIPS.
- Complete the missing half: Augmenting aggregation filtering with diversification for graph convolutional networks. arXiv preprint arXiv:2008.08844.
- Revisiting graph neural networks: All we have is low-pass filters. arXiv preprint arXiv:1905.09550.
- Multi-view Contrastive Graph Clustering. In NeurIPS, 2148–2159.
- Beyond Homophily: Reconstructing Structure for Graph-agnostic Clustering. In ICML.
- A critical look at the evaluation of GNNs under heterophily: are we really making progress? arXiv preprint arXiv:2302.11640.
- Deep clustering: A comprehensive survey. arXiv preprint arXiv:2210.04142.
- The graph neural network model. TNNLS, 20(1): 61–80.
- Multi-View Subspace Clustering via Structured Multi-Pathway Network. TNNLS.
- Powerful graph convolutional networks with adaptive propagation mechanism for homophily and heterophily. In AAAI, 4210–4218.
- Beyond Low-Pass Filtering: Graph Convolutional Networks With Automatic Filtering. TKDE, 35: 6687–6697.
- Multi-view graph embedding clustering network: Joint self-supervision and block diagonal representation. Neural Networks, 145: 1–9.
- Self-supervised discriminative feature learning for deep multi-view clustering. TKDE, 35(7): 7470–7482.
- Contrastive Learning under Heterophily. arXiv preprint arXiv:2303.06344.
- Graph Debiased Contrastive Learning with Joint Representation Clustering. In IJCAI, 3434–3440.
- Semantically consistent multi-view representation learning. Knowl. Based Syst., 278: 110899.
- MCoCo: Multi-level Consistency Collaborative multi-view clustering. Expert Systems with Applications, 238: 121976.
- Beyond homophily in graph neural networks: Current limitations and effective designs. In NeurIPS, 7793–7804.