Research map / Graph learning / Graph algorithms and clustering
Graph Neural Network Guided Local Search for the Traveling Salesperson Problem
Benjamin Hudson, Qingbiao Li, Matthew Malencia, Amanda Prorok
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 GNN-predicted global regret to guide GLS | TSP solver · TSP50 | optimality gap (%): 0.268% -> 0.009% | Improved |
| use GNN-predicted global regret to guide GLS | TSP solver · TSP100 | optimality gap (%): 1.534% -> 0.705% | Improved |
| add learned regret guidance to local search (ablation) | TSP solver · TSP100 | optimality gap (%): 4.169±2.046 -> 2.183±1.383 | Improved |
| add learned regret guidance to local search (ablation) | TSP solver · TSP50 | optimality gap (%): 3.357±2.602 -> 0.082±0.254 | Improved |
| add four additional input features to edge weight (ablation) | regret approximation model · TSP100 | optimality gap (%): 2.183±1.383 -> 2.228±1.421 | Hurt |
| add four additional input features to edge weight (ablation) | regret approximation model · TSP50 | optimality gap (%): 0.082±0.254 -> 0.080±0.295 | No clear effect |
| train on 20-node instances and generalize to TSP100 | GNN-guided GLS · TSP100 | optimality gap (%): 18.845% -> 2.622% | Improved |
| use regret guidance instead of directly classifying optimal-tour edges | GLS edge selection · Euclidean TSP | solution quality | Improved |
Closest papers by meaning
- Learning to Solve NP-Complete Problems: A Graph Neural Network for Decision TSP (AAAI 2019 · similarity 0.94)
- ViTSP: A Vision Language Models Guided Framework for Large-Scale Traveling Salesman Problems (ICLR 2026 · similarity 0.89)
- Destroy and Repair Using Hyper-Graphs for Routing (AAAI 2025 · similarity 0.87)
- Position: Rethinking Post-Hoc Search-Based Neural Approaches for Solving Large-Scale Traveling Salesman Problems (ICML 2024 · similarity 0.86)
- GLOP: Learning Global Partition and Local Construction for Solving Large-Scale Routing Problems in Real-Time (AAAI 2024 · similarity 0.86)
- DualOpt: A Dual Divide-and-Optimize Algorithm for the Large-scale Traveling Salesman Problem (AAAI 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.
