Research map / Probabilistic and scientific modeling / Variational and generative sampling
Metric Learning for Ordinal Data
Yuan Shi, Wenzhe Li, Fei Sha
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 |
|---|---|---|---|
| learn metric with LMNN instead of Euclidean on real-valued encoding | Real encoding · Car | kNN error: 11.4±0.7 -> 5.0±0.3 | Improved |
| learn metric with LMNN on real encoding | Real encoding · Balance | kNN error: 15.2±1.1 -> 17.6±0.9 | Hurt |
| learn metric with LMNN instead of Euclidean on binary encoding | Binary encoding · Car | kNN error: 24.0±1.4 -> 4.1±0.3 | Improved |
| ordinal binary encoding instead of 1-of-K encoding with LMNN | Binary-LMNN · Nursery | kNN error: 2.3±0.1 -> 1.9±0.1 | Improved |
| ordinal binary encoding instead of 1-of-K encoding with LMNN | Binary-LMNN · Balance | kNN error: 17.8±1.2 -> 13.4±0.8 | Improved |
| jointly learn thresholds and metric instead of thresholds only (fixed identity metric) (ablation) | Ord-LMNN-Uni · Balance | kNN error: 14.5±0.7 -> 6.8±0.5 | Improved |
| jointly learn thresholds and metric instead of thresholds only (ablation) | Ord-LMNN-Uni · Car | kNN error: 4.5±0.4 -> 3.8±0.3 | Improved |
| jointly learn thresholds and metric instead of thresholds only (ablation) | Ord-LMNN-Uni · Vehicle | kNN error: 27.7±0.8 -> 25.4±0.8 | Improved |
| Beta-rectangular distribution for delta_d instead of uniform (ablation) | Ord-LMNN · Nursery | kNN error: 1.8±0.1 -> 1.6±0.1 | Improved |
| Beta-rectangular distribution for delta_d instead of uniform (ablation) | Ord-LMNN · Hayes-Roth | kNN error: 20.5±1.3 -> 18.6±1.0 | Improved |
| RecBeta versus Beta distribution for delta_d (ablation) | Ord-LMNN · Car | kNN error: 3.7±0.3 -> 3.4±0.3 | Improved |
| Ord-LMNN versus best baseline Binary-Ord-LMNN | Ord-LMNN-Beta · Hayes-Roth | kNN error: 15.4±1.2 -> 18.6±1.0 | Hurt |
Closest papers by meaning
- An Ordinal Data Clustering Algorithm with Automated Distance Learning (AAAI 2020 · similarity 0.88)
- Learnable Weighting of Intra-attribute Distances for Categorical Data Clustering with Nominal and Ordinal Attributes (TPAMI 2021 · similarity 0.86)
- Finite Sample Prediction and Recovery Bounds for Ordinal Embedding (NeurIPS 2016 · similarity 0.85)
- Robust Ordinal Embedding from Contaminated Relative Comparisons (AAAI 2019 · similarity 0.85)
- Soft Labels for Ordinal Regression (CVPR 2019 · similarity 0.85)
- SLACE: A Monotone and Balance-Sensitive Loss Function for Ordinal Regression (AAAI 2025 · similarity 0.83)
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.
