Research map / Learning theory and algorithms / Prediction and decision losses
Meta-Learning for Relative Density-Ratio Estimation
Atsutoshi Kumagai, Tomoharu Iwata, Yasuhiro Fujiwara
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 latent vectors for datasets (ablation) | Ours · relative DRE avg | test squared error: -0.814 -> -0.825 | Improved |
| add adaptation to support instances with closed-form solution (ablation) | Ours · relative DRE avg | test squared error: -0.790 -> -0.825 | Improved |
| add closed-form adaptation instead of fine-tuning linear weights (ablation) | Ours · relative DRE avg | test squared error: -0.693 -> -0.825 | Improved |
| add latent vectors for datasets (ablation) | Ours · dataset comparison avg | AUC: 94.99 -> 95.72 | Improved |
| add adaptation to support instances (ablation) | Ours · dataset comparison avg | AUC: 91.60 -> 95.72 | Improved |
| add closed-form adaptation instead of fine-tuning linear weights (ablation) | Ours · dataset comparison avg | AUC: 88.71 -> 95.72 | Improved |
| add latent vectors for datasets (ablation) | Ours · outlier detection avg | AUC: 74.70 -> 75.27 | Improved |
| add adaptation to support instances (ablation) | Ours · outlier detection avg | AUC: 74.49 -> 75.27 | Improved |
| add latent vectors for datasets (ablation) | Ours · IoT outlier detection | AUC: 97.63 -> 96.94 | Hurt |
| add adaptation to support instances (ablation) | Ours · Landmine outlier detection | AUC: 68.83 -> 65.31 | Hurt |
| use relative DRE (alpha=0.5) instead of DRE (alpha=0) (ablation) | Ours · dataset comparison avg | AUC: 67.80 -> 95.72 | Improved |
| make relative parameter alpha trainable (ablation) | Ours · dataset comparison avg | AUC: 95.72 -> 95.78 | No clear effect |
Closest papers by meaning
- Overcoming Saturation in Density Ratio Estimation by Iterated Regularization (ICML 2024 · similarity 0.86)
- Classification Diffusion Models: Revitalizing Density Ratio Estimation (NeurIPS 2024 · similarity 0.85)
- Binary Losses for Density Ratio Estimation (ICLR 2025 · similarity 0.85)
- Projection Pursuit Density Ratio Estimation (ICML 2025 · similarity 0.85)
- Density Ratio Estimation with Conditional Probability Paths (ICML 2025 · similarity 0.84)
- Density Ratio Estimation with Doubly Strong Robustness (ICML 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.
