Research map / Learning theory and algorithms / Prediction and decision losses
Random Rotation Ensembles
Rico Blaser, Piotr Fryzlewicz
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 random rotation of feature space per tree | random forest · iris (ntree=5000, mtry=1) | % error: 5.227 -> 4.385 | Improved |
| add random rotation of feature space per tree | random forest · iris (ntree=50, mtry=4) | % error: 4.963 -> 4.077 | Improved |
| add random rotation of feature space per tree | extra trees · iris (ntree=5000, ncut=2) | % error: 5.175 -> 4.003 | Improved |
| add random rotation of feature space per tree | extra trees · iris (ntree>=50) | % error: 21.7% average improvement | Improved |
| add random rotation per sub-forest of 50 trees | gradient boosting machine · iris | % error: 5.063 -> 3.831 | Improved |
| add random rotation per sub-forest, rank-only inputs | gradient boosting machine · iris | % error: 5.063 -> 4.385 | Improved |
| add random rotation to random forest on UCI datasets | random forest · UCI datasets | average rank: 7.20 -> 6.38 | Improved |
| add random rotation to extra trees on UCI datasets | extra trees · UCI datasets | average rank: 7.06 -> 7.26 | Hurt |
| use rotation forests instead of random rotation random forest | random rotation random forest · UCI datasets | average rank: 6.38 -> 10.38 | Hurt |
| use quantile/complex scaling vs simple scaling vs ranking before rotation (ablation) | random rotation ensembles · UCI datasets | average rank: 7.50, 7.65, 7.81 | No clear effect |
| add random rotation to random forest, correlation between base learners | random forest · iris (mtry=1) | average margin correlation: 0.32 -> 0.22 | Improved |
| add random rotation to random forest, correlation between base learners | random forest · iris (mtry=4) | average margin correlation: 0.61 -> 0.39 | Improved |
Closest papers by meaning
- Subsampled Ensemble Can Improve Generalization Tail Exponentially (NeurIPS 2025 · similarity 0.85)
- Free Lunch in the Forest: Functionally-Identical Pruning of Boosted Tree Ensembles (AAAI 2025 · similarity 0.85)
- Diverse Prototypical Ensembles Improve Robustness to Subpopulation Shift (ICML 2025 · similarity 0.85)
- MEPSI: An MDL-Based Ensemble Pruning Approach with Structural Information (AAAI 2024 · similarity 0.85)
- Generalization Performance of Ensemble Clustering: From Theory to Algorithm (ICML 2025 · similarity 0.84)
- Sharpness-diversity tradeoff: improving flat ensembles with SharpBalance (NeurIPS 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.
