Research map / Learning theory and algorithms / Prediction and decision losses
Geometry- and Accuracy-Preserving Random Forest Proximities
Jake S. Rhodes, Adele Cutler, Kevin R. Moon
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 |
|---|---|---|---|
| replace original proximity with RF-GAP | proximity-weighted prediction · 24 UCI datasets | train/test error slope: 2.434 -> 1.040 | Improved |
| replace OOB proximity with RF-GAP | proximity-weighted prediction · 24 UCI datasets | train/test error slope: 1.543 -> 1.040 | Improved |
| replace PBK with RF-GAP | proximity-weighted prediction · 24 UCI datasets | train/test error slope: 2.287 -> 1.040 | Improved |
| replace RFProxIH with RF-GAP | proximity-weighted prediction · 24 UCI datasets | train/test error slope: 2.147 -> 1.040 | Improved |
| replace original proximity with RF-GAP | random forest imputation · UCI, 5% MCAR | average rank: 2.55 -> 1.10 | Improved |
| replace original proximity with RF-GAP | random forest imputation · UCI, 75% MCAR | average rank: 2.35 -> 1.65 | Improved |
| replace OOB proximity with RF-GAP | random forest imputation · UCI, 50% MCAR | average rank: 2.40 -> 1.35 | Improved |
| replace RFProxIH with RF-GAP | random forest imputation · UCI, 25% MCAR | average rank: 3.33 -> 1.15 | Improved |
| increase missing rate from 5% to 75% | RF-GAP imputation · UCI MCAR | average rank: 1.10 -> 1.65 | Hurt |
| replace original proximity with RF-GAP | proximity-weighted prediction vs RF · Diabetes | mismatch with RF predictions: 0.052 -> 0.001 | Improved |
| prune trees (minimum node size up to 50) | RF-GAP classification · classification datasets | match with RF predictions | No clear effect |
| replace original proximity with RF-GAP for MDS visualization | MDS on proximities · Sonar | visual class separation | Improved |
Closest papers by meaning
- Comparison-Based Random Forests (ICML 2018 · similarity 0.87)
- Joints in Random Forests (NeurIPS 2020 · similarity 0.87)
- Mixed-curvature decision trees and random forests (ICML 2025 · similarity 0.86)
- Randomization as Regularization: A Degrees of Freedom Explanation for Random Forest Success (JMLR 2020 · similarity 0.86)
- Minimax Rates for High-Dimensional Random Tessellation Forests (JMLR 2024 · similarity 0.86)
- DynFrs: An Efficient Framework for Machine Unlearning in Random Forest (ICLR 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.
