Research map / Probabilistic and scientific modeling / Variational and generative sampling
Community detection in sparse latent space models
Fengnan Gao, Zongming Ma, Hongsong Yuan
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 one round of local refinement after spectral initialization (ablation) | SpecLoRe · Caltech | misclustering error: 21.69% -> 17.46% | Improved |
| add one round of local refinement after spectral initialization (ablation) | SpecLoRe · Political Blog | misclustering error: 5.32% -> 4.66% | Improved |
| add one round of local refinement after spectral initialization (ablation) | SpecLoRe · Company network | misclustering error: 5.41% -> 2.70% | Improved |
| increase refinement rounds from 1 to 10 (ablation) | SpecLoRe · Caltech | misclustering error: 17.46% -> 14.58% | Improved |
| increase refinement rounds from 1 to 10 (ablation) | SpecLoRe · Company network | misclustering error: 2.70% -> 1.35% | Improved |
| increase refinement rounds from 1 to 10 (ablation) | SpecLoRe · High school | misclustering error: 0.61% -> 0.61% | No clear effect |
| use SpecLoRe R=10 instead of LSCD | SpecLoRe · Political Blog | runtime (s): 43.31 -> 0.97 | Improved |
| use SpecLoRe R=10 instead of LSCD | SpecLoRe · Caltech | misclustering error: 18.14% -> 14.58% | Improved |
| use SpecLoRe R=10 instead of LSCD | SpecLoRe · Simulation Spec 1 (tau=0.75) | misclustering error: 8.03% -> 8.20% | No clear effect |
| decrease latent position spread tau from 0.75 to 0.25 (ablation) | SpecLoRe R=10 · Simulation Spec 1 | misclustering error: 8.20% -> 0.83% | Improved |
| make network sparser (alpha from -2.14 to -2.83) (ablation) | SpecLoRe R=10 · Simulation Spec 3 | misclustering error: 2.23% -> 4.88% | Hurt |
Closest papers by meaning
- Analysis of spectral clustering algorithms for community detection: the general bipartite setting (JMLR 2019 · similarity 0.91)
- Contextual Stochastic Block Model: Sharp Thresholds and Contiguity (JMLR 2023 · similarity 0.91)
- Contextual Stochastic Block Models (NeurIPS 2018 · similarity 0.90)
- A Unified Framework for Spectral Clustering in Sparse Graphs (JMLR 2021 · similarity 0.89)
- Strong Consistency, Graph Laplacians, and the Stochastic Block Model (JMLR 2021 · similarity 0.89)
- Revisiting Instance-Optimal Cluster Recovery in the Labeled Stochastic Block Model (ICML 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.
