Traveling Salesperson Problem on TSPLIB Real-world instances 1.0
0.001Optimality Gap (%)AM + LCP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| AM + LCPSample width (M)=2560, Instance=berlin52, Protocol=fixed2021.10 | 0.001 | 7,550 | 13 | |
| DRL-2optNumber of iterations (I)=2000, Instance=eil512021.10 | 0.0023 | 427 | 460 | |
| AM + LCPSample width (M)=2560, Instance=eil51, Protocol=fixed2021.10 | 0.0073 | 429 | 13 | |
| DRL-2optNumber of iterations (I)=2000, Instance=st702021.10 | 0.0074 | 680 | 540 | |
| AM + LCPSample width (M)=2560, Instance=st70, Protocol=fixed2021.10 | 0.0074 | 680 | 13 | |
| AMSample width (M)=40000, Instance=eil512021.10 | 0.0211 | 435 | 13 | |
| AMSample width (M)=40000, Instance=st702021.10 | 0.0218 | 690 | 23 | |
| AM + LCPInstance=avg. gap2021.10 | 0.0514 | — | — | |
| DRL-2optNumber of iterations (I)=2000, Instance=berlin522021.10 | 0.0573 | 7,974 | 460 | |
| DRL-2optInstance=avg. gap2021.10 | 0.0763 | — | — | |
| AMInstance=avg. gap2021.10 | 0.099 | — | — | |
| AMSample width (M)=40000, Instance=berlin522021.10 | 0.1486 | 8,663 | 14 |