Empowering Targeted Neighborhood Search via Hyper Tour for Large-Scale TSP
About
Traveling Salesman Problem (TSP) is a classic NP-hard problem that has garnered significant attention from both academia and industry. While neural-based methods have shown promise for solving TSPs, they still face challenges in scaling to larger instances, particularly in memory constraints associated with global heatmaps, edge weights, or access matrices, as well as in generating high-quality initial solutions and insufficient global guidance for efficiently navigating vast search spaces. To address these challenges, we propose a Hyper Tour Guided Neighborhood Search (HyperNS) method for large-scale TSP instances. Inspired by the ``clustering first, route second" strategy, our approach initially divides the TSP instance into clusters using a sparse heatmap graph and abstracts them as supernodes, followed by the generation of a hyper tour to guide both the initialization and optimization processes. This method reduces the search space by focusing on edges relevant to the hyper tour, leading to more efficient and effective optimization. Experimental results on both synthetic and real-world datasets demonstrate that our approach outperforms existing neural-based methods, particularly in handling larger-scale instances, offering a significant reduction in the gap to the optimal solution.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Traveling Salesman Problem | TSP 1K (test) | Length23.2 | 45 | |
| Traveling Salesman Problem | Uniform-TSP1000 | Optimality Gap0.3 | 18 | |
| Traveling Salesman Problem | TSP5K generated | Tour Length51.06 | 13 | |
| Traveling Salesman Problem | TSP10K generated | Solution Length72.44 | 12 | |
| Traveling Salesman Problem | TSP20K generated | Tour Length102.4 | 8 | |
| Traveling Salesman Problem | TSP50K generated | Solution Length162.4 | 6 | |
| Traveling Salesperson Problem | TSP uniform distribution 5K | Gap (%)0.31 | 4 | |
| Traveling Salesperson Problem | TSP uniform distribution 10K | Solution Gap (%)0.93 | 4 | |
| Traveling Salesperson Problem | TSP Clustered distribution 1K | Solution Gap (%)0.41 | 4 | |
| Traveling Salesperson Problem | TSP Clustered distribution 5K | Solution Gap (%)0.55 | 4 |