Research map / Graph learning / GNN architectures
Stability of Graph Scattering Transforms
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 GST instead of GFT for graph representation | Graph signals on a 100-node small-world graph · Small-world synthetic representation stability | relative representation error: up to 3 orders of magnitude less relative representation error | Helped |
| Use GST instead of GFT | Jane Austen authorship attribution · Authorship attribution | classification accuracy | No clear effect |
| Use GST instead of GFT | Jane Austen authorship attribution · Authorship attribution under graph-estimation uncertainty | classification accuracy variability | Helped |
| Use Tight Hann GST instead of GFT | Graph source localization · Facebook 234-node subnetwork | classification accuracy | Helped |
| Use Diffusion scattering instead of GFT | Graph source localization · Facebook 234-node subnetwork | classification accuracy | Helped |
| Use Monic Cubic GST instead of GFT | Graph source localization · Facebook 234-node subnetwork | classification accuracy | No clear effect |
| Use GST instead of GFT | Graph source localization · Facebook 234-node subnetwork under edge dropping | classification accuracy variability | Mixed |
| Use Tight Hann wavelets rather than Monic Cubic wavelets | GST · Small-world synthetic representation stability | relative representation error | Helped |
| Use GST instead of GFT | Graph representations under controlled graph perturbations · Small-world synthetic representation stability | relative representation error: up to 3 orders of magnitude less relative representation error | Helped |
Closest papers by meaning
- Unsupervised Parameter-free Simplicial Representation Learning with Scattering Transforms (ICML 2024 · similarity 0.88)
- A General Graph Spectral Wavelet Convolution via Chebyshev Order Decomposition (ICML 2025 · similarity 0.87)
- WaveNet: Tackling Non-stationary Graph Signals via Graph Spectral Wavelets (AAAI 2024 · similarity 0.87)
- HoloNets: Spectral Convolutions do extend to Directed Graphs (ICLR 2024 · similarity 0.86)
- Exploitation of a Latent Mechanism in Graph Contrastive Learning: Representation Scattering (NeurIPS 2024 · similarity 0.85)
- Unitary Convolutions for Learning on Graphs and Groups (NeurIPS 2024 · similarity 0.85)
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.