Research map / Learning theory and algorithms / Sample complexity
Linking losses for density ratio and class-probability estimation
Read the paperCheck 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 |
|---|---|---|---|
| Use general CPE losses (logistic, square, exponential, square hinge) for density ratio estimation via covariate shift reweighting | density ratio estimation for covariate shift · poly and amazon covariate shift | NMSE / pairwise disagreement | Helped |
| Square loss used for density ratio estimation | covariate shift adaptation · amazon | pairwise disagreement: 0.1310 vs 0.1582 uniform | Helped |
| Exponential loss used for density ratio estimation | covariate shift adaptation · amazon | pairwise disagreement: 0.2021 vs 0.1582 uniform | Hurt |
| LSIF loss used for ranking the best instead of logistic regression | bipartite ranking (ranking the best) · german, magic, news20-forsale | AP: over logistic on all three datasets | Helped |
| Use p-classification loss (p=4) instead of logistic loss for ranking the best | bipartite ranking · news20-forsale | AP: 0.2817 vs 0.1487 | Helped |
| Canonical LSIF loss (uniform density-ratio weighting) for covariate shift | density ratio estimation · poly and amazon | NMSE / pairwise disagreement | Mixed |
| Unconstrained LSIF with post-hoc score truncation (uLSIF) vs constrained LSIF (ablation) | density ratio estimation · amazon | pairwise disagreement: 0.1370 vs 0.1500 | Helped |
Closest papers by meaning
- Binary Losses for Density Ratio Estimation (ICLR 2025 · similarity 0.91)
- Bounds on $L_p$ Errors in Density Ratio Estimation via $f$-Divergence Loss Functions (ICLR 2025 · similarity 0.90)
- Density Ratio Estimation with Doubly Strong Robustness (ICML 2024 · similarity 0.87)
- Density Ratio Estimation with Conditional Probability Paths (ICML 2025 · similarity 0.86)
- Classification Diffusion Models: Revitalizing Density Ratio Estimation (NeurIPS 2024 · similarity 0.85)
- Projection Pursuit Density Ratio Estimation (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.