Research map / Label-efficient and robust learning / Domain adaptation and generalization
Learn on Source, Refine on Target: A Model Transfer Learning Framework with Random Forests
Noam Segev, Maayan Harel, Shie Mannor, Koby Crammer, Ran El‐Yaniv
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 |
|---|---|---|---|
| Combine STRUT and SER forests into MIX (union, 100 trees) | SER · letter | error rate (%): 18.9 -> 16.7 | Improved |
| Combine STRUT and SER forests into MIX | STRUT · letter | error rate (%): 21.0 -> 16.7 | Improved |
| Combine STRUT and SER forests into MIX | SER · digits | error rate (%): 2.9 -> 3.8 | Hurt |
| Combine STRUT and SER forests into MIX | STRUT · Activity (median) | error rate (%): 14.6 -> 13.8 | Improved |
| Combine STRUT and SER into MIX when STRUT is poor | SER · inversion (1%) | error rate (%): 58.5 -> 44.1 | Mixed |
| Combine STRUT and SER into MIX when SER is poor | SER · inversion (5%) | error rate (%): 36.5 -> 22.2 | Improved |
| Combine STRUT and SER into MIX | STRUT · high-res (5%) | error rate (%): 28.3 -> 21.8 | Improved |
| Use 100 trees instead of 50 for STRUT and SER (ablation) | STRUT and SER · all datasets | error rate (%) | No clear effect |
| MIX (STRUT+SER) vs STRUT alone on moving source | STRUT · Toy moving source | test error: 6.1 -> 6.4 | Mixed |
| MIX vs SER alone on mixed boxes | SER · Toy mixed boxes | test error: 6.6 -> 5.5 | Improved |
| Use MIX instead of TradaBoost (instance transfer) | TradaBoost · letter | error rate (%): 41.7 -> 16.7 | Improved |
| Use MIX instead of TrBagg (instance transfer) | TrBagg · letter | error rate (%): 29.3 -> 16.7 | Improved |
Closest papers by meaning
- Learning Decision Trees and Forests with Algorithmic Recourse (ICML 2024 · similarity 0.85)
- DynFrs: An Efficient Framework for Machine Unlearning in Random Forest (ICLR 2025 · similarity 0.85)
- Mixed-Sample SGD: an End-to-end Analysis of Supervised Transfer Learning (NeurIPS 2025 · similarity 0.84)
- Transfer Learning of Linear Regression with Multiple Pretrained Models: Benefiting from More Pretrained Models via Overparameterization Debiasing (NeurIPS 2026 · similarity 0.84)
- Trained Random Forests Completely Reveal your Dataset (ICML 2024 · similarity 0.84)
- Wasserstein Transfer Learning (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.
