Research map / Learning theory and algorithms / Prediction and decision losses
Born-Again Tree Ensembles
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 the Random Forest with an exactly faithful minimum-size Born-Again Tree | Random Forest classifier · Six datasets | accuracy: 0.804 -> 0.804 | No clear effect |
| Post-prune branches whose two sides do not both contain training samples (ablation) | Born-Again Tree · Six datasets (average) | number of leaves: 567.0 -> 32.8 | Helped |
| Post-prune branches whose two sides do not both contain training samples (ablation) | Born-Again Tree · Six datasets (average) | accuracy: 0.804 -> 0.803 | No clear effect |
| Post-prune branches whose two sides do not both contain training samples (ablation) | Born-Again Tree · Breast-Cancer | accuracy: 0.953 -> 0.946 | Hurt |
| Post-prune branches whose two sides do not both contain training samples (ablation) | Born-Again Tree · Pima-Diabetes | accuracy: 0.746 -> 0.750 | Helped |
| Optimize leaf count instead of depth (ablation) | Born-Again Tree · Six datasets (average) | number of leaves: 567.0 -> 227.1 | Helped |
| Optimize leaf count instead of depth (ablation) | Born-Again Tree · Six datasets (average) | depth: 9.3 -> 11.8 | Hurt |
| Prioritize depth, then minimize leaf count (ablation) | Born-Again Tree · Six datasets (average) | number of leaves: 567.0 -> 266.4 | Helped |
| Use sampled-cell information-gain splits and integer-programming homogeneity checks rather than exact size optimization | Born-Again Tree · Six datasets (average) | runtime (s): 6.74 -> 0.28 | Helped |
| Use sampled-cell information-gain splits and integer-programming homogeneity checks rather than exact size optimization | Born-Again Tree · Six datasets (average) | depth: +22.53% | Hurt |
| Use sampled-cell information-gain splits and integer-programming homogeneity checks rather than exact size optimization | Born-Again Tree · Six datasets (average) | number of leaves: +20.10% | Hurt |
Closest papers by meaning
- Multistage Defer Trees for Hybrid Interpretability: If at First You Can't Succeed, Tree Again (NeurIPS 2026 · similarity 0.88)
- Compressing tree ensembles through Level-wise Optimization and Pruning (ICML 2025 · similarity 0.87)
- Near-Optimal Decision Trees in a SPLIT Second (ICML 2025 · similarity 0.86)
- Free Lunch in the Forest: Functionally-Identical Pruning of Boosted Tree Ensembles (AAAI 2025 · similarity 0.85)
- Learning Decision Trees and Forests with Algorithmic Recourse (ICML 2024 · similarity 0.85)
- Optimal Classification Trees for Continuous Feature Data Using Dynamic Programming with Branch-and-Bound (AAAI 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.