SCIENCE CHINA Information Sciences, Volume 62 , Issue 11 : 219103(2019) https://doi.org/10.1007/s11432-018-9829-5

On the neutrality of two symmetric TSP solvers toward instance specification

More info
  • ReceivedOct 9, 2018
  • AcceptedMar 15, 2019
  • PublishedSep 18, 2019


There is no abstract available for this article.


Appendix A presents the descriptive characterizations for rotated and reflected files.


[1] University of Waterloo. Concorde TSP solver. 2015. https://uwaterloo.ca/. Google Scholar

[2] Lin S, Kernighan B W. An Effective Heuristic Algorithm for the Traveling-Salesman Problem. Operations Res, 1973, 21: 498-516 CrossRef Google Scholar

[3] Neos Server. Concorde. 2019. https://neos-server.org/neos/. Google Scholar

[4] University of Bacau. Romanian TSP instance. 2017. http://cadredidactice.ub.ro/. Google Scholar

[5] University of Heidelberg. TSPLIB. 2018. https://www.uni-heidelberg.de/. Google Scholar

[6] University of Waterloo. VLSI TSP datasets. 2013. https://uwaterloo.ca/. Google Scholar

[7] Rutgers University. DIMACS generator for random TSP instances. 2013. http://archive.dimacs.rutgers.edu/. Google Scholar

[8] Mu Z, Dubois-Lacoste J, Hoos H H, et al. On the empirical scaling of running time for finding optimal solutions to the TSP. J Heuristics, 2018, 6: 879-898. Google Scholar

  • Figure 1

    (Color online) Result charts on rotated (ROT) vs. reflected (REF) files for Concorde and Lin–Kernighan. protectłinebreak (a) Variation of the Concorde solving time, grouped by instance; (b) variation of the Lin–Kernighan solution length, grouped by instance.