Research map / Learning theory and algorithms / Prediction and decision losses
Iterative Connecting Probability Estimation for Networks
Yichen Qin, Linhan Yu, Yang Li
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 |
|---|---|---|---|
| ICE vs NS on complicated graphons 4 | ICE · Graphon 4 | RMSE x100: 7.25 -> 6.09 | Improved |
| ICE vs NS on Graphon 5 | ICE · Graphon 5 | RMSE x100: 6.85 -> 4.19 | Improved |
| ICE vs NS on Graphon 8 | ICE · Graphon 8 | RMSE x100: 3.88 -> 3.30 | Improved |
| ICE vs NS on simple Graphon 1 (low rank) | ICE · Graphon 1 | RMSE x100: 2.55 -> 2.23 | No clear effect |
| ICE vs USVT on low-rank Graphon 1 | ICE · Graphon 1 | RMSE x100: 1.62 -> 2.23 | Hurt |
| two-stage (Cit=0.2, Cest=1) instead of fixed C = 0.2 (ablation) | ICE neighborhood selection · simulated network, P350,. | estimate smoothness | Improved |
| two-stage (Cit=0.2, Cest=1) instead of fixed large C = 1 (ablation) | ICE neighborhood selection · simulated network, P350,. | local structure | Improved |
| force Cit = Cest versus two-stage strategy (ablation) | ICE · simulated network | RMSE x100 | Improved |
| smaller Cit with fixed Cest (ablation) | ICE · simulated network | RMSE x100 | Improved |
| increase Cest beyond about 1 (ablation) | ICE · simulated network | RMSE x100 | Hurt |
| run iterations versus no iteration from initial estimate (ablation) | ICE · simulated network | RMSE x100 | Improved |
| ICE vs NS and USVT on link prediction | ICE · brain projectome link prediction | AUC | Improved |
Closest papers by meaning
- Estimating the Number of Induced Subgraphs from Incomplete Data and Neighborhood Queries (AAAI 2021 · similarity 0.88)
- Transfer Learning on Edge Connecting Probability Estimation Under Graphon Model (NeurIPS 2025 · similarity 0.86)
- Edge Sampling Using Local Network Information (JMLR 2021 · similarity 0.86)
- Latent distance estimation for random geometric graphs (NeurIPS 2019 · similarity 0.86)
- Node Regression on Latent Position Random Graphs via Local Averaging (JMLR 2026 · similarity 0.86)
- Navigable Graphs for High-Dimensional Nearest Neighbor Search: Constructions and Limits (NeurIPS 2024 · similarity 0.84)
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.
