Research map / Learning theory and algorithms / Combinatorial and search optimization
Efficient Active Search for Combinatorial Optimization Problems
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 |
|---|---|---|---|
| Update a subset of POMO parameters through EAS-Emb | POMO · TSP n=100 | Gap: 0.074% -> 0.063% | Helped |
| Update a subset of POMO parameters through EAS-Lay | POMO · CVRP n=100 | Gap: 0.17% -> -0.23% | Helped |
| Update static node embeddings during search | L2D · JSSP 10 × 10 | Gap: 8.0% -> 3.7% | Helped |
| Use EAS-Emb for larger-than-training instances | POMO · TSP n=200 | Gap: 1.416% -> 0.400% | Helped |
| Use EAS-Tab without retuning alpha on larger instances | POMO · CVRP n=200 | Gap: 5.35% -> 11.8% | Hurt |
| Add the imitation-learning loss to EAS-Emb or EAS-Lay (ablation) | POMO-based EAS · TSP and CVRP | solution quality | Helped |
| Add the imitation-learning loss to EAS-Emb or EAS-Lay (ablation) | L2D-based EAS · JSSP | solution quality | No clear effect |
| Add the imitation-learning loss to original active search (ablation) | Active search · TSP | Gap: 0.046% -> 0.052% | No clear effect |
| Add the imitation-learning loss to original active search (ablation) | Active search · CVRP | Gap: -0.070% -> -0.073% | No clear effect |
| Add the imitation-learning loss to original active search (ablation) | Active search · JSSP | Gap: 5.80% -> 5.78% | No clear effect |
| Reduce sigma so the lookup table has less influence (ablation) | EAS-Tab · TSP, CVRP, and JSSP | solution quality | Hurt |
Closest papers by meaning
- Improving Generalization of Neural Combinatorial Optimization for Vehicle Routing Problems via Test-Time Projection Learning (NeurIPS 2025 · similarity 0.87)
- Preference Optimization for Combinatorial Optimization Problems (ICML 2025 · similarity 0.87)
- Neural Solver Selection for Combinatorial Optimization (ICML 2025 · similarity 0.86)
- Regularized Large Neighborhood Search (NeurIPS 2026 · similarity 0.86)
- Scaling Combinatorial Optimization Neural Improvement Heuristics with Online Search and Adaptation (AAAI 2025 · similarity 0.86)
- Learning to Insert for Constructive Neural Vehicle Routing Solver (NeurIPS 2025 · similarity 0.86)
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.