Research map / Graph learning / Graph self-supervised learning
DeepNC: Deep Generative Network Completion
Cong Tran, Won-Yong Shin, Andreas Spitz, Michael Gertz
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 |
|---|---|---|---|
| add EM iterations on top of DeepNC-L (ablation) | DeepNC-L · Protein (RN) | graph edit distance: 0.3015 -> 0.2616 | Improved |
| add EM iterations on top of DeepNC-L (ablation) | DeepNC-L · LFR (FF) | graph edit distance: 0.2801 -> 0.2612 | Improved |
| add EM iterations on top of DeepNC-L (ablation) | DeepNC-L · B-A (RN, 30% nodes observed) | graph edit distance: 0.2635 -> 0.2611 | No clear effect |
| use DeepNC-EM instead of vGraphRNN | DeepNC-EM · Ego-CiteSeer (FF) | graph edit distance: 0.5416 -> 0.3241 | Improved |
| use DeepNC-EM instead of KronEM | DeepNC-EM · B-A (RN) | graph edit distance: 0.3921 -> 0.1782 | Improved |
| use DeepNC-EM instead of EvoGraph | DeepNC-EM · B-A (RN) | graph edit distance: 0.5612 -> 0.1782 | Improved |
| use forest fire instead of random node sampling | EvoGraph · LFR | graph edit distance: 0.5126 -> 0.4512 | Improved |
| use forest fire instead of random node sampling | DeepNC-EM · LFR | graph edit distance: 0.2793 -> 0.2612 | No clear effect |
| reduce observed nodes from 70% to 30% | KronEM · Protein | graph edit distance: 0.4565 -> 0.9674 | Hurt |
| reduce observed nodes from 70% to 30% | DeepNC-EM · Protein | graph edit distance: 0.2616 -> 0.3244 | No clear effect |
| make 10% of training graph nodes missing | DeepNC-L · Ego-Facebook | graph edit distance | Hurt |
Closest papers by meaning
- DepthLGP: Learning Embeddings of Out-of-Sample Nodes in Dynamic Networks (AAAI 2018 · similarity 0.88)
- Bag of Tricks for Training Deeper Graph Neural Networks: A Comprehensive Benchmark Study (TPAMI 2022 · similarity 0.87)
- GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models (ICML 2018 · similarity 0.87)
- DRGCN: Dynamic Evolving Initial Residual for Deep Graph Convolutional Networks (AAAI 2023 · similarity 0.86)
- Neural Common Neighbor with Completion for Link Prediction (ICLR 2024 · similarity 0.86)
- Propagate and Inject: Revisiting Propagation-Based Feature Imputation for Graphs with Partially Observed Features (ICML 2025 · 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.
