Research map / Learning theory and algorithms / Prediction and decision losses
Uncoupled Regression from Pairwise Comparison Data
Liyuan Xu, Junya Honda, Gang Niu, Masashi Sugiyama
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 |
|---|---|---|---|
| Use RA risk approximation instead of SVMRank | uncoupled linear regression · benchmark datasets, n_rank=5,000 | MSE: 110.3(29.5) -> 29.5(6.9); 8575.9(883.1) -> 3087.3(256.3); 62.1(7.6) -> 23.7(2.0); 322.9(45.8) -> 111.7(13.2); 372.2(34.8 | Improved |
| Use TT target transformation instead of SVMRank | uncoupled linear regression · benchmark datasets, n_rank=5,000 | MSE: 110.3(29.5) -> 22.5(6.2); 8575.9(883.1) -> 3127.3(278.8); 62.1(7.6) -> 22.7(2.2); 322.9(45.8) -> 139.1(17.9); 372.2(34.8 | Improved |
| Use proposed pairwise-comparison regression rather than supervised LR | uncoupled linear regression · benchmark datasets, n_rank=5,000 | MSE: housing: 24.5(5.0) -> 22.5(6.2); diabetes: 3041.9(219.8) -> 3087.3(256.3) / 3127.3(278.8); airfoil: 23.3(2.2) -> 22.7(2. | Mixed |
| Use TT rather than RA | uncoupled regression · synthetic data with sufficient pairwise comparisons | MSE | Improved |
| Use TT rather than RA with few pairwise comparisons | uncoupled regression · synthetic data with small pairwise-comparison sample | MSE | Hurt |
| Use TT rather than RA | uncoupled regression · housing, n_rank=5,000 | MSE: 29.5(6.9) -> 22.5(6.2) | Mixed |
Closest papers by meaning
- Linear Regression With Unmatched Data: A Deconvolution Perspective (JMLR 2024 · similarity 0.86)
- Nonparametric Regression with Comparisons: Escaping the Curse of Dimensionality with Ordinal Information (ICML 2018 · similarity 0.85)
- Mediated Uncoupled Learning: Learning Functions without Direct Input-output Correspondences (ICML 2021 · similarity 0.85)
- Regression with Multi-Expert Deferral (ICML 2024 · similarity 0.82)
- Out-Of-Domain Unlabeled Data Improves Generalization (ICLR 2024 · similarity 0.82)
- Piecewise Constant and Linear Regression Trees: An Optimal Dynamic Programming Approach (ICML 2024 · similarity 0.82)
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.
