Research map / Graph learning / GNN architectures
Self-Interpretable Subgraph Neural Network with Deep Reinforcement Walk Exploration
Jianming Huang, Hiroyuki Kasai
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 walk-exploration MDP instead of subgraph-generation MDP (ablation) | RWE-SGNN · MUTAG | accuracy (%): 92.57 -> 97.31 | Improved |
| use walk-exploration MDP instead of subgraph-generation MDP (ablation) | RWE-SGNN · NCI109 | accuracy (%): 71.99 -> 83.93 | Improved |
| use walk-exploration MDP instead of subgraph-generation MDP (ablation) | RWE-SGNN · PTC-MR | accuracy (%): 75.60 -> 75.55 | No clear effect |
| use walk-exploration MDP instead of subgraph-generation MDP (ablation) | RWE-SGNN · MUTAG | substructure generation time (ms): 124.09 -> 44.81 | Improved |
| use walk-exploration MDP instead of subgraph-generation MDP (ablation) | RWE-SGNN · NCI109 | substructure generation time (ms): 134.58 -> 51.97 | Improved |
| RWE-SGNN vs GraphSNN | RWE-SGNN · PROTEINS | accuracy (%): 78.21 -> 76.81 | Hurt |
| RWE-SGNN vs GraphSNN | RWE-SGNN · MUTAG | accuracy (%): 94.40 -> 97.31 | Improved |
| self-interpretable RWE-SGNN vs post-hoc GFlowExplainer | RWE-SGNN · BA-2motifs | edge AUC: 0.854 -> 0.918 | Improved |
| increase sample number per graph (ablation) | RWE-SGNN · NCI109 | training stability | Improved |
| vary trajectory length L from 4 to 32 (ablation) | RWE-SGNN · NCI109 | accuracy | No clear effect |
Closest papers by meaning
- Self-Interpretable Graph Learning with Sufficient and Necessary Explanations (AAAI 2024 · similarity 0.90)
- EiG-Search: Generating Edge-Induced Subgraphs for GNN Explanation in Linear Time (ICML 2024 · similarity 0.90)
- How Interpretable Are Interpretable Graph Neural Networks? (ICML 2024 · similarity 0.89)
- Generating In-Distribution Proxy Graphs for Explaining Graph Neural Networks (ICML 2024 · similarity 0.89)
- Translating Subgraphs to Nodes Makes Simple GNNs Strong and Efficient for Subgraph Representation Learning (ICML 2024 · similarity 0.89)
- Efficient Subgraph GNNs by Learning Effective Selection Policies (ICLR 2024 · 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.
