Research map / Learning theory and algorithms / Combinatorial and search optimization
Efficient Meta Neural Heuristic for Multi-Objective Combinatorial Optimization
Jinbiao Chen, Jiahai Wang, Zizhen Zhang, Zhiguang Cao, Te Ye, Siyuan Chen
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 |
|---|---|---|---|
| Use symmetric sampling only (EMNH-S) instead of scaled symmetric sampling (EMNH) (ablation) | EMNH · Bi-CVRP n=20 | HV: 0.4158 -> 0.4299 | Hurt |
| Use random weight sampling (EMNH-R) instead of scaled symmetric sampling (ablation) | EMNH · Bi-CVRP n=20 | HV: 0.3774 -> 0.4299 | Hurt |
| Random sampling vs symmetric sampling on balanced Bi-TSP-1 (ablation) | EMNH · Bi-TSP-1 n=50 | HV: 0.6339 -> 0.6362 | Hurt |
| Add scaled weight sampling to PMOCO (PMOCO-S) (ablation) | PMOCO · Bi-CVRP n=100 | HV: 0.3913 -> 0.4042 | Improved |
| Add scaled sampling to PMOCO on Bi-CVRP n=20 (ablation) | PMOCO · Bi-CVRP n=20 | HV: 0.4267 -> 0.4274 | Improved |
| Add instance augmentation to EMNH | EMNH · Bi-CVRP n=50 | HV: 0.4098 -> 0.4106 | Improved |
| Add instance augmentation to EMNH | EMNH · Tri-TSP-1 n=100 | HV: 0.4866 -> 0.4973 | Improved |
| EMNH (meta-learning, accelerated training) instead of MDRL | MDRL · Bi-KP n=50 | HV: 0.3530 -> 0.3561 | Improved |
| EMNH vs MDRL on easy balanced instance | MDRL · Bi-TSP-1 n=100 | HV: 0.6969 -> 0.6969 | No clear effect |
| EMNH vs PMOCO | PMOCO · Bi-CVRP n=100 | HV: 0.3913 -> 0.4072 | Improved |
| Accelerated training with multi-task architecture reuse | MDRL · training time | speed-up: 1/N of MDRL training time | Improved |
| EMNH-Aug vs WS-LKH on larger MOTSP | EMNH-Aug · Bi-TSP-1 n=100 | HV: 0.7023 vs 0.7090 | Hurt |
Closest papers by meaning
- Neural Solver Selection for Combinatorial Optimization (ICML 2025 · similarity 0.89)
- Improving Generalization of Neural Combinatorial Optimization for Vehicle Routing Problems via Test-Time Projection Learning (NeurIPS 2025 · similarity 0.88)
- Graph-Supported Dynamic Algorithm Configuration for Multi-Objective Combinatorial Optimization (ICML 2025 · similarity 0.86)
- Memory-Enhanced Neural Solvers for Routing Problems (NeurIPS 2025 · similarity 0.86)
- Learning to Insert for Constructive Neural Vehicle Routing Solver (NeurIPS 2025 · similarity 0.86)
- MTL-KD: Multi-Task Learning Via Knowledge Distillation for Generalizable 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.
