Research map / Learning theory and algorithms / Clustering
Subspace Clustering through Sub-Clusters
Weiwei Li, Jan Hannig, Sayan Mukherjee
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 |
|---|---|---|---|
| cluster through sub-clusters built from the whole dataset instead of subset points only (SBSC vs STSC) | sampling-based subspace clustering · Zipcode (N=9298, K=10) | accuracy-sub (%): 60.86 -> 72.04 | Improved |
| add sub-clusters to direction-search sampling (SBSC-DSC vs SDSC) | sampling-based subspace clustering · Extended Yale B (N=2432, K=38) | accuracy-sub (%): 60.78 -> 62.76 | Improved |
| sub-clusters with high-correlation neighbors vs STSC | sampling-based subspace clustering · MNIST (N=70000, K=10) | accuracy (%): 30.2 -> 95.74 | Improved |
| add bagging over 6 runs to SBSC (ablation) | SBSC · MNIST (N=70000, K=10) | accuracy (%): 95.74 -> 97.15 | Improved |
| add bagging over 6 runs to SBSC (ablation) | SBSC · MNIST (N=70000, K=10) | runtime (sec.): 38 -> 246 | Hurt |
| add bagging over 6 runs to SBSC (ablation) | SBSC · MNIST (N=70000, K=10) | accuracy-sub (%): 96.44 -> 95.25 | Hurt |
| add bagging over 6 runs to STSC (ablation) | STSC · MNIST (N=70000, K=10) | accuracy (%): 30.2 -> 40.12 | Improved |
| use SBSC on small dataset with many clusters rather than DSC | SBSC · Extended Yale B (N=2432, K=38) | accuracy (%): 91.69 -> 26.53 | Hurt |
| replace high-correlation neighbor search with direction search (SBSC-DSC vs SBSC) (ablation) | SBSC · Extended Yale B (N=2432, K=38) | accuracy (%): 26.53 -> 60.46 | Improved |
| replace high-correlation neighbor search with direction search (SBSC-DSC vs SBSC) (ablation) | SBSC · Zipcode (N=9298, K=10) | accuracy (%): 69.4 -> 60.84 | Hurt |
| increase number of data points N | SBSC (synthetic data) · synthetic | accuracy | Improved |
Closest papers by meaning
- Selective Sampling-based Scalable Sparse Subspace Clustering (NeurIPS 2019 · similarity 0.92)
- Convergence and Recovery Guarantees of the K-Subspaces Method for Subspace Clustering (ICML 2022 · similarity 0.90)
- Subspace Clustering via Good Neighbors (TPAMI 2019 · similarity 0.90)
- Subspace Structure-Aware Spectral Clustering for Robust Subspace Clustering (ICCV 2019 · similarity 0.90)
- Stochastic Sparse Subspace Clustering (CVPR 2020 · similarity 0.89)
- Theory of Spectral Method for Union of Subspaces-Based Random Geometry Graph (ICML 2021 · similarity 0.89)
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.
