Research map / Learning theory and algorithms / Combinatorial and search optimization
Community detection using fast low-cardinality semidefinite programming
Po-Wei Wang, J. Zico Kolter
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 |
|---|---|---|---|
| Replace Leiden's greedy local move with Locale embeddings and rounding | Leiden · DBLP, Amazon, IMDB, Youtube, Live Journal (1 iteration) | modularity: DBLP: 0.8206 -> 0.8273; Amazon: 0.9252 -> 0.9273; IMDB: 0.7046 -> 0.7054; Youtube: 0.7254 -> 0.7295; Live Journal: 0.757 | Mixed |
| Replace Leiden's greedy local move with Locale embeddings and rounding | Leiden · DBLP, Amazon, IMDB, Youtube, Live Journal (10 iterations) | modularity: DBLP: 0.8387 -> 0.8397; Amazon: 0.9341 -> 0.9344; IMDB: 0.7069 -> 0.7070; Youtube: 0.7328 -> 0.7355; Live Journal: 0.773 | Improved |
| Use Locale for two update rounds instead of greedy local moves | Greedy local move procedure · DBLP, Amazon, IMDB, Youtube, Live Journal | modularity: DBLP: 0.5898 -> 0.6692; Amazon: 0.6758 -> 0.7430; IMDB: 0.6580 -> 0.6697; Youtube: 0.6165 -> 0.6294; Live Journal: 0.665 | Mixed |
| Run Locale updates to full update instead of two rounds (ablation) | Locale · DBLP, Amazon, IMDB, Youtube, Live Journal | modularity: DBLP: 0.6692 -> 0.8160; Amazon: 0.7430 -> 0.9154; IMDB: 0.6697 -> 0.6852; Youtube: 0.6294 -> 0.7115; Live Journal: 0.654 | Improved |
| Increase cardinality to k=8 for the SDP relaxation (ablation) | Locale · Small verifiable networks | difference to SDP optimum: 10^{-4} | Improved |
| Increase cardinality to k=n (ablation) | Locale · Small verifiable networks | SDP optimality | Improved |
| Use Locale rather than SCS to reach the SDP optimum | SDP relaxation of modularity · zachary, polbook, football | speedup: 193x faster | Improved |
Closest papers by meaning
- Exact Clustering of Weighted Graphs via Semidefinite Programming (JMLR 2019 · similarity 0.86)
- Clustering with Semidefinite Programming and Fixed Point Iteration (JMLR 2022 · similarity 0.86)
- Achieving Optimal Misclassification Proportion in Stochastic Block Models (JMLR 2017 · similarity 0.86)
- Exact Community Recovery under Side Information: Optimality of Spectral Algorithms (ICLR 2025 · similarity 0.84)
- A Near-Linear Time Approximation Algorithm for Beyond-Worst-Case Graph Clustering (ICML 2024 · similarity 0.84)
- On the Robustness of Spectral Algorithms for Semirandom Stochastic Block Models (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.
