Research map / Learning theory and algorithms / Clustering
Robust Discriminative Clustering with Sparse Regularizers
Nicolas Flammarion, Balamurugan Palaniappan, Francis Bach
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 the l1 sparse regularizer to discriminative clustering (ablation) | discriminative clustering · Heart | cluster error: 0.61 -> 0.52 | Improved |
| add the l1 sparse regularizer to discriminative clustering (ablation) | discriminative clustering · DNA (1 vs 2,3) | cluster error: 0.83 -> 0.75 | Improved |
| add the l1 sparse regularizer to discriminative clustering (ablation) | discriminative clustering · Sonar | cluster error: 0.95 -> 0.92 | Improved |
| add the l1 sparse regularizer to discriminative clustering (ablation) | discriminative clustering · a1a | cluster error: 0.75 -> 0.74 | Improved |
| add the l1 sparse regularizer to discriminative clustering (ablation) | discriminative clustering · Diabetes, Breast-cancer, Australian, Liver-disorder, w1a | cluster error: 0.88 -> 0.88 | No clear effect |
| use sparse discriminative clustering instead of alternating optimization | clustering · Breast-cancer | cluster error: 0.48 -> 0.15 | Improved |
| use sparse discriminative clustering instead of K-means | clustering · w1a | cluster error: 0.16 -> 0.11 | Improved |
| use sparse discriminative clustering instead of K-means | clustering · Heart | cluster error: 0.91 -> 0.52 | Improved |
| add the l1 regularizer (non-zero lambda) with noisy dimensions (ablation) | FISTA discriminative clustering · synthetic data with noise variables | clustering error | Improved |
| tighten FISTA stopping criterion from 10^-2/log(d) to 10^-3/log(d) (ablation) | FISTA solver · synthetic runtime experiments | accuracy vs runtime | Mixed |
| use cvx solver instead of FISTA with lambda>0 | discriminative clustering · synthetic sparse problems | clustering error | Improved |
| replace cvx by FISTA for large n | optimization · synthetic runtime, n>=1000 | scalability: n >= 1000 | Improved |
Closest papers by meaning
- Detecting Meaningful Clusters From High-Dimensional Data: A Strongly Consistent Sparse Center-Based Clustering Approach (TPAMI 2020 · similarity 0.90)
- Hyperplane Clustering via Dual Principal Component Pursuit (ICML 2017 · similarity 0.89)
- Noisy Sparse Subspace Clustering (JMLR 2016 · similarity 0.88)
- Improved Learning-augmented Algorithms for k-means and k-medians Clustering (ICLR 2023 · similarity 0.88)
- Simple and Scalable Sparse k-means Clustering via Feature Ranking (NeurIPS 2020 · similarity 0.88)
- Probabilistic Dimensionality Reduction via Structure Learning (TPAMI 2017 · similarity 0.88)
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.
