Research map / Graph learning / Graph algorithms and clustering
Generalized Laplacian Eigenmaps
Hao Zhu, Piotr Koniusz
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 trace difference (COLES-S2GC) with LogDet rank-difference (GLEN-S2GC) | S2GC embedding · Cora (5 labels) | accuracy (%): 76.5±2.6 -> 78.2±2.4 | Improved |
| replace trace difference (COLES-S2GC) with LogDet in S2GC | S2GC embedding · Pubmed (5 labels) | accuracy (%): 66.0±5.2 -> 70.6±3.9 | Improved |
| use GLEN loss with GCN backbone instead of COLES | GCN backbone · Cora Full (5 labels) | accuracy (%): 47.3±1.5 -> 52.7±1.5 | Improved |
| GLEN-GCN versus semi-supervised GCN with 5 labels per class | GCN · Cora (5 labels) | accuracy (%): 67.5±4.8 -> 77.5±2.6 | Improved |
| GLEN-GCN versus semi-supervised GCN with 20 labels per class | GCN · Cora (20 labels) | accuracy (%): 79.4±1.6 -> 82.7±1.2 | Improved |
| GLEN-S2GC versus unsupervised S2GC | S2GC · Cora (5 labels) | accuracy (%): 71.4±4.4 -> 78.2±2.4 | Improved |
| GLEN-S2GC versus PCA projection of S2GC | S2GC · Cora (5 labels) | accuracy (%): 72.1±3.8 -> 78.2±2.4 | Improved |
| GLEN-S2GC versus random projection of S2GC | S2GC · Cora Full (5 labels) | accuracy (%): 31.5±1.4 -> 53.0±1.5 | Improved |
| GLEN-S2GC versus COLES-S2GC on standard splits | S2GC · Cora standard split | accuracy (%): 81.9 -> 85.10 | Improved |
| GLEN-S2GC versus COLES-S2GC on Ogbn-arxiv | S2GC · Ogbn-arxiv | test accuracy (%): 72.48±0.25 -> 72.67±0.26 | Improved |
| GLEN-S2GC clustering versus S2GC | S2GC · Cora clustering | accuracy (%): 68.96 -> 71.01 | Improved |
| LogDet surrogate instead of nuclear norm (ablation) | GLEN · Cora Full (5 labels) | accuracy (%): 50.8±1.4 -> 53.0±1.5 | Improved |
Closest papers by meaning
- Contrastive Laplacian Eigenmaps (NeurIPS 2021 · similarity 0.93)
- Exploitation of a Latent Mechanism in Graph Contrastive Learning: Representation Scattering (NeurIPS 2024 · similarity 0.88)
- Does GCL Need a Large Number of Negative Samples? Enhancing Graph Contrastive Learning with Effective and Efficient Negative Sampling (AAAI 2025 · similarity 0.87)
- Spectral Graph Sparsification Preserves Representation Geometry in Graph Neural Networks (NeurIPS 2026 · similarity 0.87)
- StructComp: Substituting propagation with Structural Compression in Training Graph Contrastive Learning (ICLR 2024 · similarity 0.87)
- S3GCL: Spectral, Swift, Spatial Graph Contrastive Learning (ICML 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.
