Research map / Learning theory and algorithms / Clustering
Scalable Exemplar-based Subspace Clustering on Class-Imbalanced Data
Chong You, Chi Li, Daniel P. Robinson, Rene Vidal
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 |
|---|---|---|---|
| select exemplars with FFS instead of at random (ablation) | ESC-Rand · GTSRB (14 classes, imbalanced) | clustering accuracy: 89.7 -> 93.0 | Improved |
| select exemplars with FFS instead of at random (ablation) | ESC-Rand · GTSRB (14 classes, imbalanced) | average F-score: 75.5 -> 85.3 | Improved |
| select exemplars with FFS instead of at random (variance) (ablation) | ESC-Rand · GTSRB (14 classes, imbalanced) | accuracy std: 1.6 -> 1.3 | Improved |
| express points with exemplars (ESC-FFS) instead of all points (SSC) | SSC · GTSRB (14 classes, imbalanced) | clustering accuracy: 92.4 -> 93.0 | Improved |
| express points with exemplars (ESC-FFS) instead of all points (SSC) | SSC · GTSRB (14 classes, imbalanced) | average F-score: 82.3 -> 85.3 | Improved |
| use exemplars (ESC-FFS) instead of SSC for speed | SSC · GTSRB (14 classes, imbalanced) | running time (sec.): 52.2 -> 25.2 | Improved |
| use FFS instead of random exemplars on EMNIST (ablation) | ESC-Rand · EMNIST (26 classes) | clustering accuracy | Improved |
| use ESC-FFS instead of SSC with 200 exemplars | SSC · EMNIST (26 classes) | clustering accuracy | Improved |
| select subset with FFS instead of random sampling, SRC classifier | Rand · Extended Yale B subset selection | SRC accuracy: 84.7 -> 91.4 | Improved |
| select subset with FFS instead of random sampling, SVM classifier | Rand · Extended Yale B subset selection | SVM accuracy: 83.7 -> 91.0 | Improved |
| select subset with FFS instead of K-medoids, NN classifier | K-medoids · Extended Yale B subset selection | NN accuracy: 75.5 -> 67.5 | Hurt |
| select subset with FFS instead of kDPP, SRC classifier | kDPP · Extended Yale B subset selection | SRC accuracy: 88.3 -> 91.4 | Improved |
Closest papers by meaning
- Self-Representation Based Unsupervised Exemplar Selection in a Union of Subspaces (TPAMI 2020 · similarity 0.90)
- Leveraging Union of Subspace Structure to Improve Constrained Clustering (ICML 2017 · similarity 0.88)
- Exclusivity-Consistency Regularized Multi-View Subspace Clustering (CVPR 2017 · similarity 0.87)
- Subspace Clustering through Sub-Clusters (JMLR 2021 · similarity 0.87)
- Exploring a Principled Framework for Deep Subspace Clustering (ICLR 2025 · similarity 0.85)
- Heterogeneous Sufficient Dimension Reduction and Subspace Clustering (ICML 2025 · similarity 0.85)
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.
