Research map / Graph learning / Graph self-supervised learning
Unsupervised Graph Embedding via Adaptive Graph Learning
Rui Zhang, Yunxing Zhang, Chengjun Lu, Xuelong Li
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace GCN autoencoder (GAE) with adaptive graph learning plus Laplacian term (BAGE) | GAE · Cora clustering (0% missing edges) | ACC (%): 59.53 -> 72.57 | Improved |
| Use BAGE instead of strongest baseline ARGE | ARGE · Cora clustering (50% missing edges) | ACC (%): 55.21 -> 64.22 | Improved |
| Use VBAGE instead of ARVGE | ARVGE · Cora clustering (0% missing edges) | ACC (%): 67.66 -> 73.11 | Improved |
| Use BAGE instead of ARGE | ARGE · Citeseer clustering (0% missing edges) | ACC (%): 58.67 -> 64.64 | Improved |
| Use BAGE instead of ARGE at heavy edge loss | ARGE · Citeseer clustering (50% missing edges) | ACC (%): 45.42 -> 39.91 | Hurt |
| Use BAGE instead of LGAE at heavy edge loss | LGAE · Citeseer clustering (50% missing edges) | ACC (%): 46.83 -> 39.91 | Hurt |
| Increase missing edge ratio from 0% to 50% for BAGE (ablation) | BAGE · Cora clustering | ACC (%): 72.57 -> 64.22 | Hurt |
| Increase missing edge ratio from 0% to 50% for GAE (ablation) | GAE · Cora clustering | ACC (%): 59.53 -> 48.77 | Hurt |
| Use adaptive adjacency instead of kNN-initialised graph (BAGE vs GAE) | GAE · COIL node classification | F1 (%): 93.86 -> 97.50 | Improved |
| Use BAGE instead of GAE on ATT | GAE · ATT node classification | F1 (%): 78.48 -> 86.59 | Improved |
| Use BAGE instead of GAE on PIE | GAE · PIE node classification | F1 (%): 87.69 -> 95.29 | Improved |
| Use VBAGE instead of BAGE on IMM | BAGE · IMM node classification | F1 (%): 36.71 -> 25.81 | Hurt |
Closest papers by meaning
- Permutation-Invariant Variational Autoencoder for Graph-Level Representation Learning (NeurIPS 2021 · similarity 0.90)
- GraphZoom: A Multi-level Spectral Approach for Accurate and Scalable Graph Embedding (ICLR 2020 · similarity 0.90)
- Graph Representation Learning via Ladder Gamma Variational Autoencoders (AAAI 2020 · similarity 0.89)
- Dual Low-Rank Graph Autoencoder for Semantic and Topological Networks (AAAI 2023 · similarity 0.88)
- HC-GAE: The Hierarchical Cluster-based Graph Auto-Encoder for Graph Representation Learning (NeurIPS 2024 · similarity 0.88)
- GraphCroc: Cross-Correlation Autoencoder for Graph Structural Reconstruction (NeurIPS 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
