Research map / Learning theory and algorithms / Prediction and decision losses
Tree ensemble optimization
Improve, prune and interpret random forests and boosted tree ensembles.
77 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Prediction and decision losses, keyword cluster “trees · decision tree · forests”
State History Tried Checklist Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
12 papers in 2023–24 and 19 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic grew, from 2.2% to 3.3%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Evasion and Hardening of Tree Ensemble Classifiers (ICML 2016 · 101 citations)
- 2017 · LightGBM: A Highly Efficient Gradient Boosting Decision Tree (NeurIPS 2017 · 9,433 citations)
- 2018 · When do random forests fail? (NeurIPS 2018 · 42 citations)
- 2019 · A Debiased MDI Feature Importance Measure for Random Forests (NeurIPS 2019 · 80 citations)
- 2020 · Regional Tree Regularization for Interpretability in Deep Neural Networks (AAAI 2020 · 36 citations)
- 2021 · Are Neural Rankers still Outperformed by Gradient Boosted Decision Trees? (ICLR 2021 · 41 citations)
- 2022 · Trading Complexity for Sparsity in Random Forest Explanations (AAAI 2022 · 26 citations)
- 2023 · When Do Neural Nets Outperform Boosted Trees on Tabular Data? (NeurIPS 2023 · 72 citations)
- 2024 · MEPSI: An MDL-Based Ensemble Pruning Approach with Structural Information (AAAI 2024 · 3 citations)
- 2025 · Free Lunch in the Forest: Functionally-Identical Pruning of Boosted Tree Ensembles (AAAI 2025 · 2 citations)
- 2026 · CLARITree: Cholesky and Lookahead Accelerations for Regression with Interpretable Piecewise Linear Trees (ICML 2026 · 0 citations)
Earliest papers here
- Evasion and Hardening of Tree Ensemble Classifiers (ICML 2016 · 101 citations)
- Learning Online Smooth Predictors for Realtime Camera Planning Using Recurrent Decision Trees (CVPR 2016 · 53 citations)
- Online Learning With Bayesian Classification Trees (CVPR 2016 · 3 citations)
What has been tried
64 papers here report 587 measured changes (a component added, removed or swapped, compared with a baseline): helped 64%, no clear effect 11%, hurt 17%, mixed 9%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| Replace the exact ∥·∥0 pruning objective with the approximate ∥·∥1 objective | FICO · pruning time | 456 -> 30 | Helped | Free Lunch in the Forest: Functionally-Identical Pruning of Boosted Tree Ensembles (AAAI 2025 · 2 citations) |
| Use minimal angle maximization instead of precomputed hyperspherical points | Binary tree (depth 8 edges, 10 dimensions) · D_wc | 18.42 -> 1.635 | Helped | Low-distortion and GPU-compatible Tree Embeddings in Hyperbolic Space (ICML 2025 · 1 citation) |
| Increase λ from 1 to 10 | Synthetic classification d=20 · accuracy | 52.56 (1.02) -> 57.42 (0.47) | Helped | Invariant Random Forest: Tree-Based Model Solution for OOD Generalization (AAAI 2024 · 0 citations) |
| Increase maximum tree depth | Gaussian mixture classification · micro-F1 | Hurt | Fast Hyperboloid Decision Tree Algorithms (ICLR 2024 · 0 citations) | |
| Vary the ordering of hierarchical attributes | Hierarchical group collections · misclassification error | No clear effect | Multi-group Learning for Hierarchical Groups (ICML 2024 · 0 citations) | |
| Add ℓ1-discontinuity regularization | Cpuact · test RMSE | Helped | Generalized additive models via direct optimization of regularized decision stump forests (ICML 2025) |
What reviewers will expect
Datasets and baselines named in the 12 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Datasets
- California Housing · 2 of 12 papers
- FICO · 2 of 12 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 31 papers; from OpenAlex and the papers' affiliations).
- Columbia University · 3 papers
- Duke University · 2 papers
- Carnegie Mellon University · 2 papers
- Google · 2 papers
Use it
Download BibTeX of 40 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Optimal decision tree learning (75 papers)
- Compact interpretable model classes (26 papers)
Elsewhere in the topic
- Adaptive testing and experiments (201 papers)
- Active learning and subset selection (199 papers)
- Distributionally robust optimization (184 papers)
- Data attribution and recourse (160 papers)
- Theory of neural network sensitivity (143 papers)
- Ranking models from comparisons (127 papers)
- Determinantal point process inference (122 papers)
- Conformal prediction sets (116 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).