Research map / Learning theory and algorithms / Combinatorial and search optimization
SAT-based Decision Tree Learning for Large Data Sets
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 |
|---|---|---|---|
| Apply SAT-based local subtree improvement to Weka decision trees | Weka · 37 UCI and benchmark instances | average depth: 22.63 -> 13.45 | Helped |
| Apply SAT-based local subtree improvement to Weka decision trees | Weka · 37 UCI and benchmark instances | average test accuracy: 0.67 -> 0.76 | Helped |
| Apply SAT-based local subtree improvement to ITI decision trees | ITI · 37 UCI and benchmark instances | average depth: 10.75 -> 8.35 | Helped |
| Apply SAT-based local subtree improvement to ITI decision trees | ITI · 37 UCI and benchmark instances | average test accuracy: 0.77 -> 0.78 | Helped |
| Use the hybrid DT depth and DT pb encoding rather than DT depth alone (ablation) | DT-SLIM(Weka) · 37 UCI and benchmark instances | average depth: 34% lower depth | Helped |
| Use the hybrid DT depth and DT pb encoding rather than DT depth alone (ablation) | DT-SLIM(ITI) · 37 UCI and benchmark instances | average depth: 2% | Helped |
| Increase the timeout and subtree limits from d60, c60 to d300, c300 (ablation) | DT-SLIM(Weka) · 37 UCI and benchmark instances | average depth: 13.64 -> 13.45 | Helped |
| Increase the timeout and subtree limits from d300, c300 to d800, c800 (ablation) | DT-SLIM(Weka) · 37 UCI and benchmark instances | average depth: 13.45 -> 14.33 | Hurt |
| Increase the timeout and subtree limits from d60, c60 to d300, c300 (ablation) | DT-SLIM(ITI) · 37 UCI and benchmark instances | average depth: 8.55 -> 8.35 | Helped |
| Increase the timeout and subtree limits from d300, c300 to d800, c800 (ablation) | DT-SLIM(ITI) · 37 UCI and benchmark instances | average depth: 8.35 -> 8.59 | Hurt |
| Reduce the Weka tree using SAT-based local improvement | Weka · australian | depth: 53.00 -> 22.00 | Helped |
| Reduce the Weka tree using SAT-based local improvement | Weka · objectivity | test accuracy: 0.55 -> 0.78 | Helped |
Closest papers by meaning
- Branches: Efficiently Seeking Optimal Sparse Decision Trees via AO* (ICML 2025 · similarity 0.90)
- Improving Decision Trees through the Lens of Parameterized Local Search (NeurIPS 2025 · similarity 0.89)
- Witty: An Efficient Solver for Computing Minimum-Size Decision Trees (AAAI 2025 · similarity 0.89)
- Optimal Decision Tree Pruning Revisited: Algorithms and Complexity (ICML 2025 · similarity 0.88)
- Near-Optimal Decision Trees in a SPLIT Second (ICML 2025 · similarity 0.88)
- Learning Small Decision Trees with Few Outliers: A Parameterized Perspective (AAAI 2024 · similarity 0.87)
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.