Ensemble graph auto-encoders for clustering and link prediction
Graph auto-encoders are a crucial research area within graph neural networks, commonly employed for generating graph embeddings while minimizing errors in unsupervised learning. Traditional graph auto-encoders focus on reconstructing minimal graph data loss to encode neighborhood information for eac...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PeerJ Inc.
2025-01-01
|
| Series: | PeerJ Computer Science |
| Subjects: | |
| Online Access: | https://peerj.com/articles/cs-2648.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832587190401499136 |
|---|---|
| author | Chengxin Xie Jingui Huang Yongjiang Shi Hui Pang Liting Gao Xiumei Wen |
| author_facet | Chengxin Xie Jingui Huang Yongjiang Shi Hui Pang Liting Gao Xiumei Wen |
| author_sort | Chengxin Xie |
| collection | DOAJ |
| description | Graph auto-encoders are a crucial research area within graph neural networks, commonly employed for generating graph embeddings while minimizing errors in unsupervised learning. Traditional graph auto-encoders focus on reconstructing minimal graph data loss to encode neighborhood information for each node, yielding node embedding representations. However, existing graph auto-encoder models often overlook node representations and fail to capture contextual node information within the graph data, resulting in poor embedding effects. Accordingly, this study proposes the ensemble graph auto-encoders (E-GAE) model. It utilizes the ensemble random walk graph auto-encoder, the random walk graph auto-encoder of the ensemble network, and the graph attention auto-encoder to generate three node embedding matrices Z. Then, these techniques are combined using adaptive weights to reconstruct a new node embedding matrix. This method addresses the problem of low-quality embeddings. The model’s performance is evaluated using three publicly available datasets (Cora, Citeseer, and PubMed), indicating its effectiveness through multiple experiments. It achieves up to a 2.0% improvement in the link prediction task and a 9.4% enhancement in the clustering task. Our code for this work can be found at https://github.com/xcgydfjjjderg/graphautoencoder. |
| format | Article |
| id | doaj-art-04419b50a5ed495ea221f2380ff9905f |
| institution | Kabale University |
| issn | 2376-5992 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | PeerJ Inc. |
| record_format | Article |
| series | PeerJ Computer Science |
| spelling | doaj-art-04419b50a5ed495ea221f2380ff9905f2025-01-24T15:05:09ZengPeerJ Inc.PeerJ Computer Science2376-59922025-01-0111e264810.7717/peerj-cs.2648Ensemble graph auto-encoders for clustering and link predictionChengxin Xie0Jingui Huang1Yongjiang Shi2Hui Pang3Liting Gao4Xiumei Wen5Hebei University of Architecture, Zhangjiakou, ChinaHunan Normal University, ChangSha, ChinaHebei University of Architecture, Zhangjiakou, ChinaHebei University of Architecture, Zhangjiakou, ChinaHebei University of Architecture, Zhangjiakou, ChinaHebei University of Architecture, Zhangjiakou, ChinaGraph auto-encoders are a crucial research area within graph neural networks, commonly employed for generating graph embeddings while minimizing errors in unsupervised learning. Traditional graph auto-encoders focus on reconstructing minimal graph data loss to encode neighborhood information for each node, yielding node embedding representations. However, existing graph auto-encoder models often overlook node representations and fail to capture contextual node information within the graph data, resulting in poor embedding effects. Accordingly, this study proposes the ensemble graph auto-encoders (E-GAE) model. It utilizes the ensemble random walk graph auto-encoder, the random walk graph auto-encoder of the ensemble network, and the graph attention auto-encoder to generate three node embedding matrices Z. Then, these techniques are combined using adaptive weights to reconstruct a new node embedding matrix. This method addresses the problem of low-quality embeddings. The model’s performance is evaluated using three publicly available datasets (Cora, Citeseer, and PubMed), indicating its effectiveness through multiple experiments. It achieves up to a 2.0% improvement in the link prediction task and a 9.4% enhancement in the clustering task. Our code for this work can be found at https://github.com/xcgydfjjjderg/graphautoencoder.https://peerj.com/articles/cs-2648.pdfGraph auto-encodersLow embeddingEnsembleLink predictionClustering |
| spellingShingle | Chengxin Xie Jingui Huang Yongjiang Shi Hui Pang Liting Gao Xiumei Wen Ensemble graph auto-encoders for clustering and link prediction PeerJ Computer Science Graph auto-encoders Low embedding Ensemble Link prediction Clustering |
| title | Ensemble graph auto-encoders for clustering and link prediction |
| title_full | Ensemble graph auto-encoders for clustering and link prediction |
| title_fullStr | Ensemble graph auto-encoders for clustering and link prediction |
| title_full_unstemmed | Ensemble graph auto-encoders for clustering and link prediction |
| title_short | Ensemble graph auto-encoders for clustering and link prediction |
| title_sort | ensemble graph auto encoders for clustering and link prediction |
| topic | Graph auto-encoders Low embedding Ensemble Link prediction Clustering |
| url | https://peerj.com/articles/cs-2648.pdf |
| work_keys_str_mv | AT chengxinxie ensemblegraphautoencodersforclusteringandlinkprediction AT jinguihuang ensemblegraphautoencodersforclusteringandlinkprediction AT yongjiangshi ensemblegraphautoencodersforclusteringandlinkprediction AT huipang ensemblegraphautoencodersforclusteringandlinkprediction AT litinggao ensemblegraphautoencodersforclusteringandlinkprediction AT xiumeiwen ensemblegraphautoencodersforclusteringandlinkprediction |