Research map / Graph learning / GNN architectures
Reservoir-Based Graph Convolutional Networks
Mayssa Soussia, Gita Ayu Salsabila, Mohamed Ali Mahjoub, Islem Rekik
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 |
|---|---|---|---|
| make reservoir weights trainable (TRGC-Net) (ablation) | RGC-Net · MUTAG (k=1, avg) | accuracy: 0.8611 -> 0.8207 | Hurt |
| make reservoir weights trainable (TRGC-Net) (ablation) | RGC-Net · EMCI-AD (avg) | accuracy: 0.4464 -> 0.4606 (k=1); 0.5120 -> 0.4456 (k=4) | Mixed |
| make reservoir weights trainable (TRGC-Net) (ablation) | RGC-Net · PROTEINS | training time (s): 1.83 -> 10.54 | Hurt |
| make reservoir weights trainable (TRGC-Net) (ablation) | RGC-Net · DD | training time (s): 7.73 -> 38.92 | Hurt |
| replace GCN layer with reservoir graph convolution layer | GCN classifier · MUTAG (1 layer) | accuracy: 0.7500 -> 0.8726 | Improved |
| increase number of layers from 1 to 5 (ablation) | RGC-Net · MUTAG | accuracy: 0.8726 -> 0.8032 | Hurt |
| increase reservoir iterations from 1 to 5 (ablation) | RGC-Net · MUTAG (avg) | accuracy: 0.8611 -> 0.8063 | Hurt |
| increase reservoir iterations from 1 to 4 (ablation) | RGC-Net · EMCI-AD (avg) | accuracy: 0.4464 -> 0.5120 | Improved |
| lower leaky rate from 1 to 0.8 (k=1) (ablation) | RGC-Net · MUTAG | accuracy: 0.8442 -> 0.8726 | Improved |
| lower leaky rate from 1 to 0.8 (k=1) (ablation) | RGC-Net · EMCI-AD | accuracy: 0.4476 -> 0.4573 | No clear effect |
| use RGC-Net layers instead of GCN layers in graph transformer | Graph transformer · EMCI-AD | MAE: 0.6847 -> 0.1860 | Improved |
| use RGC-Net layers instead of GCN layers in graph transformer | Graph transformer · Simulated | MAE: 0.0353 -> 0.0305 | Improved |
Closest papers by meaning
- Generalizing CNNs to graphs with learnable neighborhood quantization (NeurIPS 2024 · similarity 0.90)
- L2-GCN: Layer-Wise and Learned Efficient Training of Graph Convolutional Networks (CVPR 2020 · similarity 0.90)
- Simplifying Graph Convolutional Networks (ICML 2019 · similarity 0.89)
- EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs (AAAI 2020 · similarity 0.89)
- Efficient Learning of Linear Graph Neural Networks via Node Subsampling (NeurIPS 2023 · similarity 0.89)
- On the Stability of Graph Convolutional Neural Networks: A Probabilistic Perspective (NeurIPS 2025 · similarity 0.89)
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.
