Research map / Learning theory and algorithms / Clustering
The Cluster Structure Function
Andrew R. Cohen, Paul M. B. Vitanyi
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 |
|---|---|---|---|
| use CSF curves as input to a shallow feedforward network instead of Gap Statistic curves | MNIST digit-count estimation · MNIST digit sets | accuracy: 54% -> 86% | Improved |
| use CSF instead of Gap Statistic for unsupervised K selection (K true > 1) | MNIST digit sets · MNIST digit sets | correlation r: -0.84 -> 0.93 | Improved |
| omit digit sets with K true = 1 from the correlation (ablation) | CSF unsupervised selection · MNIST digit sets | correlation | Improved |
| use FLIF compressor instead of BZIP and JPEG2000 (ablation) | NCD for MNIST · MNIST digit sets | clustering quality | Improved |
| ensemble segmentation with optimality deficiency selection instead of best single radius | cell segmentation · Cell tracking challenge (12 movies) | DET: p=5e-4 | Improved |
| ensemble segmentation instead of best single radius | cell segmentation · Cell tracking challenge (12 movies) | TRA: p=2e-3 | Improved |
| CSF instead of Gap, AIC, BIC for estimating K | synthetic 3 Gaussian clusters · Synthetic K=3 | K estimation accuracy | Improved |
Closest papers by meaning
- Compact Representation of Uncertainty in Clustering (NeurIPS 2018 · similarity 0.84)
- Hierarchical Clustering Beyond the Worst-Case (NeurIPS 2017 · similarity 0.83)
- Near-perfect Clustering Based on Recursive Binary Splitting Using Max-MMD (TPAMI 2026 · similarity 0.83)
- Gradient Based Clustering (ICML 2022 · similarity 0.82)
- Clustering by Mining Density Distributions and Splitting Manifold Structure (AAAI 2025 · similarity 0.81)
- SEC: More Accurate Clustering Algorithm via Structural Entropy (AAAI 2024 · similarity 0.81)
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.
