Research map / Learning theory and algorithms / Prediction and decision losses
Comparison-Based Random Forests
Siavash Haghiri, Damien Garreau, Ulrike Luxburg
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 label-informed supervised pivots instead of random unsupervised pivots (ablation) | CompRF · MNIST | classification error: 4.21 (± 0.05) -> 2.50 (± 0.05) | Improved |
| Use label-informed supervised pivots instead of random unsupervised pivots (ablation) | CompRF · Gisette | classification error: 3.28 (± 0.19) -> 2.48 (± 0.13) | Improved |
| Use label-informed supervised pivots instead of random unsupervised pivots (ablation) | CompRF · UCIHAR | classification error: 8.70 (± 0.32) -> 6.54 (± 0.11) | Improved |
| Use label-informed supervised pivots instead of random unsupervised pivots (ablation) | CompRF · Isolet | classification error: 6.65 (± 0.14) -> 4.43 (± 0.26) | Improved |
| Use label-informed supervised pivots instead of random unsupervised pivots (ablation) | CompRF · MUTAG, WL-subtree kernel | classification error: 14.44 (± 7.94) -> 13.89 (± 7.97) | No clear effect |
| Use label-informed supervised pivots instead of random unsupervised pivots (ablation) | CompRF · ENZYMES, WL-subtree kernel | classification error: 39.33 (± 6.49) -> 39.83 (± 5.00) | Mixed |
| Use label-informed supervised pivots instead of random unsupervised pivots (ablation) | CompRF · NCI1, WL-edge kernel | classification error: 17.91 (± 1.42) -> 18.05 (± 1.63) | Mixed |
| Use comparison-based trees rather than CART forests | CompRF supervised · MNIST | classification error: 2.90 (± 0.05) -> 2.50 (± 0.05) | Improved |
| Use comparison-based trees rather than CART forests | CompRF supervised · Gisette | classification error: 3.04 (± 0.26) -> 2.48 (± 0.13) | Improved |
| Use comparison-based trees rather than CART forests | CompRF supervised · UCIHAR | classification error: 7.47 (± 0.32) -> 6.54 (± 0.11) | Improved |
| Use comparison-based trees rather than CART forests | CompRF supervised · Isolet | classification error: 5.48 (± 0.27) -> 4.43 (± 0.26) | Improved |
| Use comparison-based trees rather than CART forests | CompRF · Boston | RMSE: 3.02 (± 0.95) -> 6.16 (± 1.00) | Hurt |
Closest papers by meaning
- Geometry- and Accuracy-Preserving Random Forest Proximities (TPAMI 2023 · similarity 0.87)
- Towards Convergence Rate Analysis of Random Forests for Classification (NeurIPS 2020 · similarity 0.87)
- Randomization as Regularization: A Degrees of Freedom Explanation for Random Forest Success (JMLR 2020 · similarity 0.86)
- Explaining the Success of AdaBoost and Random Forests as Interpolating Classifiers (JMLR 2017 · similarity 0.86)
- Hierarchical Shrinkage: Improving the accuracy and interpretability of tree-based models. (ICML 2022 · similarity 0.86)
- Optimal Weighted Random Forests (JMLR 2024 · similarity 0.86)
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.
