Skip to main content
SpringerLink
Log in

On the empirical time complexity of finding optimal solutions vs proving optimality for Euclidean TSP instances

  • Short Communication
  • Published:
Optimization Letters Aims and scope Submit manuscript

Abstract

We investigate the empirical performance of the long-standing state-of-the-art exact TSP solver Concorde on various classes of Euclidean TSP instances and show that, surprisingly, the time spent until the first optimal solution is found accounts for a large fraction of Concorde’s overall running time. This finding holds for the widely studied random uniform Euclidean (RUE) instances as well as for several other widely studied sets of Euclidean TSP instances. On RUE instances, the median fraction of Concorde’s total running time spent until an optimal solution is found ranges from 0.77 for \(n=500\) to 0.97 for \(n=3{,}500\); on TSPLIB, National and VLSI instances, we pegged it at 0.86, 0.74 and 0.61, respectively, with a tendency of even smaller values for larger instances.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Notes

  1. We further discuss the reasons for holding such perceptions in Sect. 4.

  2. Note that on larger instances, no optimal solutions were found in Concorde’s initial CLK phase.

  3. We note that for solving very large TSP instances, non-standard settings of Concorde are used, and special techniques are exploited to obtain feasible computation times on clusters of parallel machines; for details, see Chapter 16 of [3]. In fact, for the proof of optimality for the instance sw24978 (which involves the locations of cities in Sweden), it was stated explicitly that “the long Sweden computation benefited greatly form the accurate upper bounds provided by Helsgaun’s tour” [3], p. 517.

References

  1. The traveling salesman problem. Version visited last on 15 April 2014. http://www.math.uwaterloo.ca/tsp/ (2014)

  2. Applegate, D.: Personal communication (2009)

  3. Applegate, D., Bixby, R.E., Chvatal, V., Cook, W.J.: The traveling salesman problem: a computational study. Princeton University Press, Princeton (2006)

    Google Scholar 

  4. Applegate, D., Bixby, R.E., Chvatal, V., Cook, W.J.: Concorde TSP solver. Version visited last on 15 April 2014. http://www.math.uwaterloo.ca/tsp/concorde.html (2014)

  5. Beame, P., Pitassi, T.: Propositional proof complexity: past, present, and future. In: Current trends in theoretical computer science: entering the 21st century, pp. 42–70. World Scientific Publishing (2001)

  6. Dash, S.: Exponential lower bounds on the lengths of some classes of branch-and-cut proofs. Math. Oper. Res. 30(3), 678–700 (2005)

    Article  MathSciNet  MATH  Google Scholar 

  7. Helsgaun, K.: An effective implementation of the Lin-Kernighan traveling salesman heuristic. Eur. J. Oper. Res. 126, 106–130 (2000)

    Article  MathSciNet  MATH  Google Scholar 

  8. Helsgaun, K.: General k-opt submoves for the Lin-Kernighan TSP heuristic. Math. Program. Comput. 1(2–3), 119–163 (2009)

    Article  MathSciNet  MATH  Google Scholar 

  9. Hoos, H.H., Stützle, T.: On the empirical scaling of run-time for finding optimal solutions to the traveling salesman problem. Eur. J. Oper. Res. 238(1), 87–94 (2014)

    Article  Google Scholar 

  10. Lawler, E.L., Lenstra, J.K., Rinnooy Kan., A.H.G, Shmoys, D.B.: The traveling salesman problem. Wiley, Chichester (1985)

  11. Nagata, Y., Kobayashi, S.: A powerful genetic algorithm using edge assembly crossover for the traveling salesman problem. INFORMS J. Comput. 25(2), 346–363 (2013)

    Article  MathSciNet  Google Scholar 

  12. Pipatsrisawat, K., Darwiche, A.: On the power of clause-learning SAT solvers as resolution engines. Artif. Intell. 175, 512–525 (2011)

    Article  MathSciNet  MATH  Google Scholar 

  13. Reinelt, G.: The traveling salesman: computational solutions for TSP applications. Lecture Notes in Computer Science, vol. 840. Springer, Heidelberg, Germany (1994)

  14. Reinelt, G.: TSPLIB. Version visited last on 15 June 2012, http://www.iwr.uni-heidelberg.de/groups/comopt/software/TSPLIB95 (2012)

Download references

Acknowledgments

We gratefully acknowledge helpful input from David Mitchell on connections with complexity theory. Furthermore, we thank the anonymous reviewers for their useful comments. This work was supported by the COMEX project within the Interuniversity Attraction Poles Programme of the Belgian Science Policy Office. H. H. acknowledges support through an NSERC Discovery Grant. T. S. acknowledges support from the Belgian F. R. S.-FNRS, of which he is a senior research associate.

Author information

Authors and Affiliations

  1. Computer Science Department, University of British Columbia, Vancouver, BC, Canada

    Holger H. Hoos

  2. IRIDIA, CoDE, Université Libre de Bruxelles (ULB), Brussels, Belgium

    Thomas Stützle

Authors
  1. Holger H. Hoos
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Thomas Stützle
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Holger H. Hoos.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 164 KB)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hoos, H.H., Stützle, T. On the empirical time complexity of finding optimal solutions vs proving optimality for Euclidean TSP instances. Optim Lett 9, 1247–1254 (2015). https://doi.org/10.1007/s11590-014-0828-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11590-014-0828-5

Keywords

Search

Navigation