Research map / Label-efficient and robust learning / Recognition and long-tail
Precision-based Boosting
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 |
|---|---|---|---|
| Replace SAMME's total-error weighting with class-specific precision weighting | SAMME (PrSAMME) · connect4 | test error (%): 29.58 -> 28.14 | Helped |
| Replace SAMME's total-error weighting with class-specific precision weighting | SAMME (PrSAMME) · sensorless | test error (%): 59.64 -> 45.64 | Helped |
| Replace SAMME's total-error weighting with class-specific precision weighting | SAMME (PrSAMME) · shuttle | test error (%): 5.54 -> 0.89 | Helped |
| Replace SAMME's total-error weighting with class-specific precision weighting | SAMME (PrSAMME) · letter | test error (%): 73.06 -> 66.30 | Helped |
| Replace SAMME's total-error weighting with class-specific precision weighting | SAMME (PrSAMME) · splice | test error (%): 6.80 -> 7.18 | Hurt |
| Replace SAMME's total-error weighting with class-specific precision weighting | SAMME (PrSAMME) · iris | test error (%): 5.95 -> 7.13 | Hurt |
| Replace SAMME's total-error weighting with class-specific precision weighting | SAMME (PrSAMME) · breastt | test error (%): 40.24 -> 37.26; 31.00 -> 38.55 | Mixed |
| Replace AdaBoost's total-error weighting with class-specific precision weighting | AdaBoost (PrAdaBoost) · bankfull | test error (%): 10.50 -> 10.33 | Helped |
| Replace AdaBoost's total-error weighting with class-specific precision weighting | AdaBoost (PrAdaBoost) · magic04 | test error (%): 19.14 -> 17.98 | Helped |
| Use class-specific precision weighting instead of InfoBoost's parameters | InfoBoost · adult | test error (%): 33.99 -> 15.11 | Helped |
| Use class-specific precision weighting instead of InfoBoost's parameters | InfoBoost · promoters | test error (%): 14.23 -> 15.31 | Hurt |
| Use class-specific precision weighting instead of total-error weighting | SAMME · multi-class UCI datasets, averaged over iterations 1–100 | test error (%) | Helped |
Closest papers by meaning
- AdaBoost is not an Optimal Weak to Strong Learner (ICML 2023 · similarity 0.89)
- A Boosting-Type Convergence Result for AdaBoost.MH with Factorized Multi-Class Classifiers (NeurIPS 2024 · similarity 0.88)
- The Many Faces of Optimal Weak-to-Strong Learning (NeurIPS 2024 · similarity 0.87)
- Boosting with Multiple Sources (NeurIPS 2021 · similarity 0.86)
- CatBoost: unbiased boosting with categorical features (NeurIPS 2018 · similarity 0.86)
- Robust Minimax Boosting with Performance Guarantees (NeurIPS 2025 · 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.