Research map / Reinforcement learning / Visual and representation RL
Elite Pattern Reinforcement for Vehicle Routing Problems
Ning Li, Peng Lin, Peng Zhang, Ruichen Tian
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add elite-pattern reinforcement (EPR) (ablation) | ELG-POMO · CVRPLIB-X total | optimality gap (%): 6.42 -> 5.96 | Improved |
| add EPR (ablation) | ELG-POMO · TSPLIB total | optimality gap (%): 3.30 -> 3.13 | Improved |
| add EPR | POMO · CVRPLIB-X total | optimality gap (%): 16.08 -> 10.54 | Improved |
| add EPR | Sym-POMO · CVRPLIB-X total | optimality gap (%): 30.54 -> 24.61 | Improved |
| add EPR | Omni-POMO · CVRPLIB-X total | optimality gap (%): 6.31 -> 6.19 | Improved |
| add EPR | POMO · TSPLIB total | optimality gap (%): 6.15 -> 6.09 | No clear effect |
| add EPR | ELG-POMO · CVRPLIB-M | optimality gap (%): 3.99 -> 3.72 | Improved |
| add EPR | POMO · TSPLIB (0,200] | optimality gap (%): 2.29 -> 2.31 | Hurt |
| add path symmetry (reversed patterns) to core EPR (ablation) | ELG-POMO + EPR · CVRPLIB-X total | optimality gap (%): 6.32 -> 5.96 | Improved |
| add path symmetry (reversed patterns) to core EPR (ablation) | ELG-POMO + EPR · TSPLIB total | optimality gap (%): 3.23 -> 3.13 | Improved |
| add core EPR guidance without symmetry (ablation) | ELG-POMO · CVRPLIB-X total | optimality gap (%): 6.42 -> 6.32 | Improved |
| vary guidance proportion P (0.05, 0.1, 0.2) (ablation) | EPR-ELG-POMO · CVRPLIB-X total | optimality gap (%): 6.28, 5.96, 6.10 | No clear effect |
Closest papers by meaning
- MVMoE: Multi-Task Vehicle Routing Solver with Mixture-of-Experts (ICML 2024 · similarity 0.87)
- Memory-Enhanced Neural Solvers for Routing Problems (NeurIPS 2025 · similarity 0.87)
- Rethinking Light Decoder-based Solvers for Vehicle Routing Problems (ICLR 2025 · similarity 0.87)
- MTL-KD: Multi-Task Learning Via Knowledge Distillation for Generalizable Neural Vehicle Routing Solver (NeurIPS 2025 · similarity 0.87)
- Multi-Task Vehicle Routing Solver via Mixture of Specialized Experts under State-Decomposable MDP (NeurIPS 2025 · similarity 0.86)
- Distilling Autoregressive Models to Obtain High-Performance Non-autoregressive Solvers for Vehicle Routing Problems with Faster Inference Speed (AAAI 2024 · 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.
