Research map / Continual learning / Online and replay-based
Graph Memory Learning: Imitating Lifelong Remembering and Forgetting of Brain Networks
Jiaxing Miao, Liang Hu, Qi Zhang, Longbing Cao
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 BGML instead of strongest baseline FGNs on regular node classification | BGML · Cora regular | F1-score: 86.80 -> 88.89 | Improved |
| use BGML instead of SEM in graph memory learning (remember and forget) | BGML · Cora memory | F1-score: 86.12 -> 86.30 | Improved |
| use BGML instead of SEM in graph memory learning | BGML · CiteSeer memory | F1-score: 77.13 -> 79.28 | Improved |
| use BGML instead of GraphDelete in unlearning | BGML · Cora unlearning | F1-score: 86.37 -> 87.93 | Improved |
| use BGML instead of TACO in class-incremental learning | BGML · Cora class-incremental | F1-score: 82.26 -> 83.33 | Improved |
| add MGHPL multi-granular hierarchical progressive learning (second-level submodels support weaker first-level ones) (ablation) | BGML · Cora | F1-score | Improved |
| add ISAO so new information selects its own partition (ablation) | BGML · Cora | F1-score | Improved |
| use BGML in unlearning versus GraphEraser in forgetting speed | BGML · Cora, CiteSeer, PubMed | unlearning time | Hurt |
| use BGML instead of baselines for catastrophic forgetting | BGML · class-incremental Cora, CiteSeer, PubMed | forgetting rate | Improved |
Closest papers by meaning
- Towards Continuous Reuse of Graph Models via Holistic Memory Diversification (ICLR 2025 · similarity 0.89)
- Exploring Rationale Learning for Continual Graph Learning (AAAI 2025 · similarity 0.88)
- What Matters in Graph Class Incremental Learning? An Information Preservation Perspective (NeurIPS 2024 · similarity 0.88)
- A Topology-aware Graph Coarsening Framework for Continual Graph Learning (NeurIPS 2024 · similarity 0.87)
- A Selective Learning Method for Temporal Graph Continual Learning (ICML 2025 · similarity 0.87)
- Memorization in Graph Neural Networks (NeurIPS 2025 · 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.
