Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Generalize a Small Pre-trained Model to Arbitrarily Large TSP Instances

About

For the traveling salesman problem (TSP), the existing supervised learning based algorithms suffer seriously from the lack of generalization ability. To overcome this drawback, this paper tries to train (in supervised manner) a small-scale model, which could be repetitively used to build heat maps for TSP instances of arbitrarily large size, based on a series of techniques such as graph sampling, graph converting and heat maps merging. Furthermore, the heat maps are fed into a reinforcement learning approach (Monte Carlo tree search), to guide the search of high-quality solutions. Experimental results based on a large number of instances (with up to 10,000 vertices) show that, this new approach clearly outperforms the existing machine learning based TSP algorithms, and significantly improves the generalization ability of the trained model.

Zhang-Hua Fu, Kai-Bin Qiu, Hongyuan Zha• 2020

Related benchmarks

TaskDatasetResultRank
Traveling Salesman ProblemTSP-500 (test)
Gap2.54
85
Traveling Salesman Problem (TSP)TSP n=100 10K instances (test)--
52
Traveling Salesperson ProblemTSP-100
Solution Length7.7638
42
Traveling Salesman ProblemTSP-1000 (test)
Optimality Gap3.22
36
Traveling Salesperson ProblemTSP N=100 (test)
Optimality Gap0.04
21
Traveling Salesperson ProblemTSP N=200 (Generalization (128 instances))
Optimality Gap0.88
19
Traveling Salesman ProblemTSP-10000 (test)
Solution Length74.93
17
Traveling Salesperson ProblemTSP N=1000 Generalization (128 instances)
Optimality Gap3.22
14
Traveling Salesperson ProblemTSP N=500 Generalization (128 instances)
Optimality Gap2.54
14
Traveling Salesman ProblemTSPLIB 50-200
Drop1.14
10
Showing 10 of 10 rows

Other info

Follow for update