Research map / Learning theory and algorithms / Combinatorial and search optimization
A Scalable Two Stage Approach to Computing Optimal Decision 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 |
|---|---|---|---|
| split decision-set learning into individual rule enumeration followed by set cover | ruler · 1065-dataset benchmark suite | datasets solved: 800 vs. 578 | Helped |
| use ILP rather than MaxSAT for the set-cover phase | ruler · 1065-dataset benchmark suite, rule minimization | datasets solved: 802 vs. 734 | Helped |
| use ILP rather than MaxSAT for the set-cover phase | ruler · 1065-dataset benchmark suite, literal minimization | datasets solved: 800 vs. 669 | Helped |
| add symmetry breaking during rule enumeration (ablation) | ruler · 1065-dataset benchmark suite | datasets solved: 686 vs. 556 | Helped |
| add symmetry breaking during rule enumeration (ablation) | ruler · 1065-dataset benchmark suite | average number of rules enumerated: 19604.4 -> 563.7 | Helped |
| minimize total literals instead of number of rules | decision-set learning · datasets solved by both mds_2 and ruler ilp-l | average number of literals: 116.2 -> 62.2 | Helped |
| lexicographically minimize rules, then literals rather than only minimizing rules | mds_2 · 1065-dataset benchmark suite | datasets solved: 578 -> 398 | Hurt |
| use literal-minimizing ruler rather than opt | minimum-size decision-set learning · 1065-dataset benchmark suite | datasets solved: 351 -> 800 | Helped |
Closest papers by meaning
- SAT-based Decision Tree Learning for Large Data Sets (AAAI 2021 · similarity 0.88)
- SAT-Based Tree Decomposition with Iterative Cascading Policy Selection (AAAI 2024 · similarity 0.87)
- Learning MAX-SAT from Contextual Examples for Combinatorial Optimisation (AAAI 2020 · similarity 0.87)
- Witty: An Efficient Solver for Computing Minimum-Size Decision Trees (AAAI 2025 · similarity 0.85)
- Learning How to Cube (NeurIPS 2026 · similarity 0.84)
- HardCore Generation: Generating Hard UNSAT Problems for Data Augmentation (NeurIPS 2024 · 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.