SSHPool: The Separated Subgraph-based Hierarchical Pooling
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866929456747118592 |
|---|---|
| author | Xu, Zhuo Cui, Lixin Li, Ming Wang, Yue Lyu, Ziyu Du, Hangyuan Bai, Lu Yu, Philip S. Hancock, Edwin R. |
| author_facet | Xu, Zhuo Cui, Lixin Li, Ming Wang, Yue Lyu, Ziyu Du, Hangyuan Bai, Lu Yu, Philip S. Hancock, Edwin R. |
| contents | In this paper, we develop a novel local graph pooling method, namely the Separated Subgraph-based Hierarchical Pooling (SSHPool), for graph classification. We commence by assigning the nodes of a sample graph into different clusters, resulting in a family of separated subgraphs. We individually employ the local graph convolution units as the local structure to further compress each subgraph into a coarsened node, transforming the original graph into a coarsened graph. Since these subgraphs are separated by different clusters and the structural information cannot be propagated between them, the local convolution operation can significantly avoid the over-smoothing problem caused by message passing through edges in most existing Graph Neural Networks (GNNs). By hierarchically performing the proposed procedures on the resulting coarsened graph, the proposed SSHPool can effectively extract the hierarchical global features of the original graph structure, encapsulating rich intrinsic structural characteristics. Furthermore, we develop an end-to-end GNN framework associated with the SSHPool module for graph classification. Experimental results demonstrate the superior performance of the proposed model on real-world datasets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_16133 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | SSHPool: The Separated Subgraph-based Hierarchical Pooling Xu, Zhuo Cui, Lixin Li, Ming Wang, Yue Lyu, Ziyu Du, Hangyuan Bai, Lu Yu, Philip S. Hancock, Edwin R. Artificial Intelligence Machine Learning In this paper, we develop a novel local graph pooling method, namely the Separated Subgraph-based Hierarchical Pooling (SSHPool), for graph classification. We commence by assigning the nodes of a sample graph into different clusters, resulting in a family of separated subgraphs. We individually employ the local graph convolution units as the local structure to further compress each subgraph into a coarsened node, transforming the original graph into a coarsened graph. Since these subgraphs are separated by different clusters and the structural information cannot be propagated between them, the local convolution operation can significantly avoid the over-smoothing problem caused by message passing through edges in most existing Graph Neural Networks (GNNs). By hierarchically performing the proposed procedures on the resulting coarsened graph, the proposed SSHPool can effectively extract the hierarchical global features of the original graph structure, encapsulating rich intrinsic structural characteristics. Furthermore, we develop an end-to-end GNN framework associated with the SSHPool module for graph classification. Experimental results demonstrate the superior performance of the proposed model on real-world datasets. |
| title | SSHPool: The Separated Subgraph-based Hierarchical Pooling |
| topic | Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2403.16133 |