Research map / Graph learning / Graph algorithms and clustering
TOPOGRAPH: Topology-Preserving Graph Reduction with Adaptive Structure for Persistent Homology
Zonghao Chen, Yuncheng Jiang, Gang 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 |
|---|---|---|---|
| use TOPOGRAPH instead of PH-landmarks | persistence homology pipeline · Three-rings (30% outliers, 40% reduction, 2D) | bottleneck distance: 0.323 -> 0.154 | Improved |
| use TOPOGRAPH instead of PH-landmarks | persistence homology pipeline · Eyeglasses (10% outliers, 40% reduction, 50D) | bottleneck distance: 0.162 -> 0.046 | Improved |
| use TOPOGRAPH instead of PH-landmarks | persistence homology pipeline · 3D-Torus (30% outliers, 40% reduction, 50D) | bottleneck distance: 0.024 -> 0.094 | Hurt |
| use TOPOGRAPH instead of PH-landmarks | persistence homology pipeline · 3D-Torus (10% outliers, 40% reduction, 2D/3D) | bottleneck distance: 0.048 -> 0.064 | Hurt |
| use TOPOGRAPH instead of PH-landmarks | persistence homology pipeline · LinkedCircle (10% outliers, 40% reduction, 50D) | bottleneck distance: 0.374 -> 0.374 | No clear effect |
| use TOPOGRAPH instead of PH-landmarks | persistence homology pipeline · Three-rings (10% outliers, 40% reduction, 2D) | runtime: 20 -> 2.5 | Improved |
| use TOPOGRAPH instead of PH-landmarks | persistence homology pipeline · Three-rings (30% outliers, 40% reduction, 2D) | residual outliers: 619 -> 428 | Improved |
| use TOPOGRAPH instead of random sampling | persistence homology pipeline · Knot | total time (s): 7588 s -> 529 s | Improved |
| vary DTM neighborhood k and AkNNG max degree kmax (ablation) | TOPOGRAPH · four synthetic datasets | bottleneck distance: 0.5%, 1.0%, 1.5%, 2.0% | No clear effect |
| vary k and kmax (ablation) | TOPOGRAPH · Three-rings | bottleneck distance: 0.5%, 1.0%, 1.5%, 2.0% | Mixed |
| vary neighborhood size k=kmax between 0.5% and 2% (ablation) | TOPOGRAPH · Neural IPCs | H1 loop recovery: 0.5% and 2% | No clear effect |
| use TOPOGRAPH instead of PH-landmarks (delta 0.01 and 0.1) | persistence homology pipeline · Neural IPCs | spurious persistence features | Improved |
Closest papers by meaning
- Persistent Homology for High-dimensional Data Based on Spectral Methods (NeurIPS 2024 · similarity 0.89)
- Graphcode: Learning from multiparameter persistent homology using graph neural networks (NeurIPS 2024 · similarity 0.88)
- Topological Periodicity Test (TopPT) via Confidence Bound of Time-Delay Embeddings (NeurIPS 2026 · similarity 0.88)
- TopER: Topological Embeddings in Graph Representation Learning (NeurIPS 2025 · similarity 0.88)
- Point-Level Topological Representation Learning on Point Clouds (ICML 2025 · similarity 0.87)
- A Class of Topological Pseudodistances for Fast Comparison of Persistence Diagrams (AAAI 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.
