Research map / Learning theory and algorithms / Clustering
Closed-Form Training of Mahalanobis Distance for Supervised Clustering
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| learn the metric with the closed-form MLCA solution rather than Euclidean distance | kmeans clustering · Synthetic, equal-size clusters | Test Δloss: 3.0 -> 0.07 | Helped |
| learn a metric with MLCA rather than KISSME | supervised clustering · Synthetic, different-size clusters with noise | Test Δloss: 2.02 -> 0.09 | Helped |
| learn a metric with MLCA rather than KISSME | supervised clustering · Synthetic, different-size clusters without noise | Test Δloss: 0 -> 2.0 | No clear effect |
| use multivariate MLCA rather than univariate MLCA (ablation) | image segmentation · Horses | Rand loss: 0.29 ± 0.02 -> 0.32 ± 0.02 | Hurt |
| use multivariate MLCA rather than univariate MLCA (ablation) | image segmentation · Horses | Δloss: 1.34 ± 0.02 -> 1.20 ± 0.02 | Helped |
| use multivariate MLCA rather than univariate MLCA (ablation) | image segmentation · Oxford Flowers-17 | Δloss: 1.47 ± 0.01 -> 1.29 ± 0.01 | Helped |
| use multivariate MLCA rather than univariate MLCA (ablation) | image segmentation · Oxford Flowers-17 | Rand loss: 0.22 ± 0.01 -> 0.22 ± 0.01 | No clear effect |
| use MLCA rather than KISSME | dynamic texture segmentation · SynthDB | Rand index: 0.590 -> 0.902 | Helped |
| learn the metric with MLCA rather than the method of Teney et al. | dynamic texture segmentation · SynthDB | Rand index: 0.902 -> 0.902 | No clear effect |
Closest papers by meaning
- Deep Spectral Clustering Learning (ICML 2017 · similarity 0.88)
- Learning to Link (ICLR 2020 · similarity 0.87)
- Fairness, Semi-Supervised Learning, and More: A General Framework for Clustering with Stochastic Pairwise Constraints (AAAI 2021 · similarity 0.87)
- Supervised Hierarchical Clustering with Exponential Linkage (ICML 2019 · similarity 0.87)
- MC-GTA: Metric-Constrained Model-Based Clustering using Goodness-of-fit Tests with Autocorrelations (ICML 2024 · similarity 0.84)
- Approximation Scheme for Weighted Metric Clustering via Sherali-Adams (AAAI 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.