Research map / Graph learning / Graph self-supervised learning
Learning Graph Representations with Embedding Propagation
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 |
|---|---|---|---|
| Learn label embeddings by reconstructing each node's label representation from neighboring nodes | EP-B · BlogCatalog, POS, PPI transductive node classification | micro F1 | Mixed |
| Use label representations and their combination for attributed citation graphs | EP-B · Cora, Citeseer, Pubmed transductive node classification | accuracy | Mixed |
| Use EP-B instead of GCN | Cora transductive · Cora | accuracy: 78.05 $\pm$ 1.49 vs. 79.59 $\pm$ 2.02 | Hurt |
| Apply EP-B to nodes absent during unsupervised training, reconstructing their embeddings from neighbors | EP-B · BlogCatalog, POS, PPI inductive node classification | micro F1 | Helped |
| Apply EP-B to nodes absent during training | EP-B · Cora, Citeseer, Pubmed inductive node classification | accuracy | Helped |
| Use directed rather than undirected edge message passing | EP-B · Cora, Citeseer, Pubmed | accuracy | No clear effect |
| Limit message senders to a sampled neighbor subset of size κ=50 | EP-B · BlogCatalog | average batch loss: 20$s per epoch to less than 1$s per epoch | No clear effect |
| Reduce the sampled neighbor limit κ | EP-B · BlogCatalog | average batch loss | Hurt |
Closest papers by meaning
- InfoGraph: Unsupervised and Semi-supervised Graph-Level Representation Learning via Mutual Information Maximization (ICLR 2020 · similarity 0.89)
- Revisiting Semi-Supervised Learning with Graph Embeddings (ICML 2016 · similarity 0.89)
- Inductive and Unsupervised Representation Learning on Graph Structured Objects (ICLR 2020 · similarity 0.88)
- GraphZoom: A Multi-level Spectral Approach for Accurate and Scalable Graph Embedding (ICLR 2020 · similarity 0.87)
- On the Equivalence between Positional Node Embeddings and Structural Graph Representations (ICLR 2020 · similarity 0.87)
- Learning Representations for Hierarchies with Minimal Support (NeurIPS 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.