Skip to main content
Top
Published in: OR Spectrum 3/2021

01-09-2020 | Regular Article

Generalization of machine learning for problem reduction: a case study on travelling salesman problems

Authors: Yuan Sun, Andreas Ernst, Xiaodong Li, Jake Weiner

Published in: OR Spectrum | Issue 3/2021

Log in

Activate our intelligent search to find suitable subject content or patents.

search-config
loading …

Abstract

Combinatorial optimization plays an important role in real-world problem solving. In the big data era, the dimensionality of a combinatorial optimization problem is usually very large, which poses a significant challenge to existing solution methods. In this paper, we examine the generalization capability of a machine learning model for problem reduction on the classic travelling salesman problems (TSP). We demonstrate that our method can greedily remove decision variables from an optimization problem that are predicted not to be part of an optimal solution. More specifically, we investigate our model’s capability to generalize on test instances that have not been seen during the training phase. We consider three scenarios where training and test instances are different in terms of: (1) problem characteristics; (2) problem sizes; and (3) problem types. Our experiments show that this machine learning-based technique can generalize reasonably well over a wide range of TSP test instances with different characteristics or sizes. Since the accuracy of predicting unused variables naturally deteriorates as a test instance is further away from the training set, we observe that, even when tested on a different TSP problem variant, the machine learning model still makes useful predictions about which variables can be eliminated without significantly impacting solution quality.

Dont have a licence yet? Then find out more about our products and how to get one now:

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Appendix
Available only for authorised users
Footnotes
1
Our C++ source codes are publicly available online at https://​github.​com/​yuansuny/​tsp.
 
3
We do not remove the edges that appear in the best sample solution to guarantee that the reduced problem space contains at least one feasible solution.
 
Literature
go back to reference Applegate D, Cook W, Rohe A (2003) Chained Lin–Kernighan for large traveling salesman problems. INFORMS J Comput 15(1):82–92CrossRef Applegate D, Cook W, Rohe A (2003) Chained Lin–Kernighan for large traveling salesman problems. INFORMS J Comput 15(1):82–92CrossRef
go back to reference Applegate D, Bixby R, Chvatal V, Cook W (2006a) Concorde TSP solver Applegate D, Bixby R, Chvatal V, Cook W (2006a) Concorde TSP solver
go back to reference Applegate DL, Bixby RE, Chvatal V, Cook WJ (2006b) The traveling salesman problem: a computational study. Princeton University Press, Princeton Applegate DL, Bixby RE, Chvatal V, Cook WJ (2006b) The traveling salesman problem: a computational study. Princeton University Press, Princeton
go back to reference Balasundaram B, Butenko S, Hicks IV (2011) Clique relaxations in social network analysis: the maximum k-plex problem. Oper Res 59(1):133–142CrossRef Balasundaram B, Butenko S, Hicks IV (2011) Clique relaxations in social network analysis: the maximum k-plex problem. Oper Res 59(1):133–142CrossRef
go back to reference Bello I, Pham H, Le QV, Norouzi M, Bengio S (2016) Neural combinatorial optimization with reinforcement learning. arXiv preprint. arXiv:1611.09940 Bello I, Pham H, Le QV, Norouzi M, Bengio S (2016) Neural combinatorial optimization with reinforcement learning. arXiv preprint. arXiv:1611.09940
go back to reference Bengio Y, Lodi A, Prouvost A (2018) Machine learning for combinatorial optimization: a methodological tour d’horizon. arXiv preprint. arXiv:1811.06128 Bengio Y, Lodi A, Prouvost A (2018) Machine learning for combinatorial optimization: a methodological tour d’horizon. arXiv preprint. arXiv:1811.06128
go back to reference Blum C, Pinacho P, López-Ibáñez M, Lozano JA (2016) Construct, merge, solve & adapt a new general algorithm for combinatorial optimization. Comput Oper Res 68:75–88CrossRef Blum C, Pinacho P, López-Ibáñez M, Lozano JA (2016) Construct, merge, solve & adapt a new general algorithm for combinatorial optimization. Comput Oper Res 68:75–88CrossRef
go back to reference Boser BE, Guyon IM, Vapnik VN (1992) A training algorithm for optimal margin classifiers. In: Proceedings of the fifth annual workshop on computational learning theory. ACM, pp 144–152 Boser BE, Guyon IM, Vapnik VN (1992) A training algorithm for optimal margin classifiers. In: Proceedings of the fifth annual workshop on computational learning theory. ACM, pp 144–152
go back to reference Chang C-C, Lin C-J (2011) LIBSVM: a library for support vector machines. ACM Trans Intell Syst Technol 2:27:1–27:27CrossRef Chang C-C, Lin C-J (2011) LIBSVM: a library for support vector machines. ACM Trans Intell Syst Technol 2:27:1–27:27CrossRef
go back to reference Chen X, Tian Y (2019) Learning to perform local rewriting for combinatorial optimization. Adv Neural Inf Process Syst 6278–6289 Chen X, Tian Y (2019) Learning to perform local rewriting for combinatorial optimization. Adv Neural Inf Process Syst 6278–6289
go back to reference Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20(3):273–297 Cortes C, Vapnik V (1995) Support-vector networks. Mach Learn 20(3):273–297
go back to reference Deudon M, Cournut P, Lacoste A, Adulyasak Y, Rousseau L-M (2018) Learning heuristics for the TSP by policy gradient. In: International conference on the integration of constraint programming, artificial intelligence, and operations research. Springer, pp 170–181 Deudon M, Cournut P, Lacoste A, Adulyasak Y, Rousseau L-M (2018) Learning heuristics for the TSP by policy gradient. In: International conference on the integration of constraint programming, artificial intelligence, and operations research. Springer, pp 170–181
go back to reference Ding J-Y, Zhang C, Shen L, Li S, Wang B, Xu Y, Song L (2019) Accelerating primal solution findings for mixed integer programs based on solution prediction. arXiv preprint. arXiv:1906.09575 Ding J-Y, Zhang C, Shen L, Li S, Wang B, Xu Y, Song L (2019) Accelerating primal solution findings for mixed integer programs based on solution prediction. arXiv preprint. arXiv:1906.09575
go back to reference Dong C, Jäger G, Richter D, Molitor P (2009) Effective tour searching for tsp by contraction of pseudo backbone edges. In: International conference on algorithmic applications in management. Springer, pp 175–187 Dong C, Jäger G, Richter D, Molitor P (2009) Effective tour searching for tsp by contraction of pseudo backbone edges. In: International conference on algorithmic applications in management. Springer, pp 175–187
go back to reference Fan R-E, Chen P-H, Lin C-J (2005) Working set selection using second order information for training support vector machines. J Mach Learn Res 6(Dec):1889–1918 Fan R-E, Chen P-H, Lin C-J (2005) Working set selection using second order information for training support vector machines. J Mach Learn Res 6(Dec):1889–1918
go back to reference Fan R-E, Chang K-W, Hsieh C-J, Wang X-R, Lin C-J (2008) LIBLINEAR: a library for large linear classification. J Mach Learn Res 9(Aug):1871–1874 Fan R-E, Chang K-W, Hsieh C-J, Wang X-R, Lin C-J (2008) LIBLINEAR: a library for large linear classification. J Mach Learn Res 9(Aug):1871–1874
go back to reference Fischer T, Merz P (2007) Reducing the size of traveling salesman problem instances by fixing edges. In: European conference on evolutionary computation in combinatorial optimization. Springer, pp 72–83 Fischer T, Merz P (2007) Reducing the size of traveling salesman problem instances by fixing edges. In: European conference on evolutionary computation in combinatorial optimization. Springer, pp 72–83
go back to reference Friggstad Z, Gollapudi S, Kollias K, Sarlos T, Swamy C, Tomkins A (2018) Orienteering algorithms for generating travel itineraries. In: Proceedings of the eleventh ACM international conference on web search and data mining. ACM, pp 180–188 Friggstad Z, Gollapudi S, Kollias K, Sarlos T, Swamy C, Tomkins A (2018) Orienteering algorithms for generating travel itineraries. In: Proceedings of the eleventh ACM international conference on web search and data mining. ACM, pp 180–188
go back to reference Gao J, Chen J, Yin M, Chen R, Wang Y (2018) An exact algorithm for maximum k-plexes in massive graphs. IJCAI 1449–1455 Gao J, Chen J, Yin M, Chen R, Wang Y (2018) An exact algorithm for maximum k-plexes in massive graphs. IJCAI 1449–1455
go back to reference Grassia M, Lauri J, Dutta S, Ajwani D (2019) Learning multi-stage sparsification for maximum clique enumeration. arXiv preprint. arXiv:1910.00517 Grassia M, Lauri J, Dutta S, Ajwani D (2019) Learning multi-stage sparsification for maximum clique enumeration. arXiv preprint. arXiv:1910.00517
go back to reference He H, Daume H III, Eisner JM (2014) Learning to search in branch and bound algorithms. Adv Neural Inf Process Syst 3293–3301 He H, Daume H III, Eisner JM (2014) Learning to search in branch and bound algorithms. Adv Neural Inf Process Syst 3293–3301
go back to reference Helsgaun K (2000) An effective implementation of the Lin–Kernighan traveling salesman heuristic. Eur J Oper Res 126(1):106–130CrossRef Helsgaun K (2000) An effective implementation of the Lin–Kernighan traveling salesman heuristic. Eur J Oper Res 126(1):106–130CrossRef
go back to reference Hougardy S, Schroeder RT (2014) Edge elimination in tsp instances. In: International workshop on graph-theoretic concepts in computer science. Springer, pp 275–286 Hougardy S, Schroeder RT (2014) Edge elimination in tsp instances. In: International workshop on graph-theoretic concepts in computer science. Springer, pp 275–286
go back to reference Jäger G, Dong C, Goldengorin B, Molitor P, Richter D (2014) A backbone based TSP heuristic for large instances. J Heuristics 20(1):107–124CrossRef Jäger G, Dong C, Goldengorin B, Molitor P, Richter D (2014) A backbone based TSP heuristic for large instances. J Heuristics 20(1):107–124CrossRef
go back to reference Johnson DS, McGeoch LA (1997) The traveling salesman problem: a case study in local optimization. Local Search Comb Optim 1(1):215–310 Johnson DS, McGeoch LA (1997) The traveling salesman problem: a case study in local optimization. Local Search Comb Optim 1(1):215–310
go back to reference Jonker R, Volgenant T (1983) Transforming asymmetric into symmetric traveling salesman problems. Oper Res Lett 2(4):161–163CrossRef Jonker R, Volgenant T (1983) Transforming asymmetric into symmetric traveling salesman problems. Oper Res Lett 2(4):161–163CrossRef
go back to reference Jonker R, Volgenant T (1984) Nonoptimal edges for the symmetric traveling salesman problem. Oper Res 32(4):837–846CrossRef Jonker R, Volgenant T (1984) Nonoptimal edges for the symmetric traveling salesman problem. Oper Res 32(4):837–846CrossRef
go back to reference Khalil E, Dai H, Zhang Y, Dilkina B, Song L (2017) Learning combinatorial optimization algorithms over graphs. Adv Neural Inf Process Syst 6348–6358 Khalil E, Dai H, Zhang Y, Dilkina B, Song L (2017) Learning combinatorial optimization algorithms over graphs. Adv Neural Inf Process Syst 6348–6358
go back to reference Kilby P, Slaney J, Walsh T et al (2005) The backbone of the travelling salesperson. IJCAI 175–180 Kilby P, Slaney J, Walsh T et al (2005) The backbone of the travelling salesperson. IJCAI 175–180
go back to reference Kool W, van Hoof H, Welling M (2019) Attention, learn to solve routing problems!. International conference on learning representations Kool W, van Hoof H, Welling M (2019) Attention, learn to solve routing problems!. International conference on learning representations
go back to reference Lauri J, Dutta S (2019) Fine-grained search space classification for hard enumeration variants of subset problems. In: Proceedings of the thirty-third AAAI conference on artificial intelligence. AAAI, pp 2314–2321 Lauri J, Dutta S (2019) Fine-grained search space classification for hard enumeration variants of subset problems. In: Proceedings of the thirty-third AAAI conference on artificial intelligence. AAAI, pp 2314–2321
go back to reference Li Z, Chen Q, Koltun V (2018) Combinatorial optimization with graph convolutional networks and guided tree search. Adv Neural Inf Process Syst 539–548 Li Z, Chen Q, Koltun V (2018) Combinatorial optimization with graph convolutional networks and guided tree search. Adv Neural Inf Process Syst 539–548
go back to reference Lin S, Kernighan BW (1973) An effective heuristic algorithm for the traveling-salesman problem. Oper Res 21(2):498–516CrossRef Lin S, Kernighan BW (1973) An effective heuristic algorithm for the traveling-salesman problem. Oper Res 21(2):498–516CrossRef
go back to reference Lin C-J, Weng RC, Keerthi SS (2008) Trust region Newton method for logistic regression. J Mach Learn Res 9(Jun):627–650 Lin C-J, Weng RC, Keerthi SS (2008) Trust region Newton method for logistic regression. J Mach Learn Res 9(Jun):627–650
go back to reference Reinelt G (1991) Tsplib—a traveling salesman problem library. ORSA J Comput 3(4):376–384CrossRef Reinelt G (1991) Tsplib—a traveling salesman problem library. ORSA J Comput 3(4):376–384CrossRef
go back to reference Sherali HD, Driscoll PJ (2002) On tightening the relaxations of Miller–Tucker–Zemlin formulations for asymmetric traveling salesman problems. Oper Res 50(4):656–669CrossRef Sherali HD, Driscoll PJ (2002) On tightening the relaxations of Miller–Tucker–Zemlin formulations for asymmetric traveling salesman problems. Oper Res 50(4):656–669CrossRef
go back to reference Smith-Miles K, van Hemert J (2011) Discovering the suitability of optimisation algorithms by learning from evolved instances. Ann Math Artif Intell 61(2):87–104CrossRef Smith-Miles K, van Hemert J (2011) Discovering the suitability of optimisation algorithms by learning from evolved instances. Ann Math Artif Intell 61(2):87–104CrossRef
go back to reference Sun Y, Li X, Ernst A (2019) Using statistical measures and machine learning for graph reduction to solve maximum weight clique problems. IEEE Trans Pattern Anal Mach Intell Sun Y, Li X, Ernst A (2019) Using statistical measures and machine learning for graph reduction to solve maximum weight clique problems. IEEE Trans Pattern Anal Mach Intell
go back to reference Vinyals O, Fortunato M, Jaitly N (2015) Pointer networks. Adv Neural Inf Process Syst 2692–2700 Vinyals O, Fortunato M, Jaitly N (2015) Pointer networks. Adv Neural Inf Process Syst 2692–2700
go back to reference Wu Q, Hao J-K (2015) A review on algorithms for maximum clique problems. Eur J Oper Res 242(3):693–709CrossRef Wu Q, Hao J-K (2015) A review on algorithms for maximum clique problems. Eur J Oper Res 242(3):693–709CrossRef
go back to reference Wu Y, Song W, Cao Z, Zhang J, Lim A (2019) Learning improvement heuristics for solving the travelling salesman problem. arXiv preprint. arXiv:1912.05784 Wu Y, Song W, Cao Z, Zhang J, Lim A (2019) Learning improvement heuristics for solving the travelling salesman problem. arXiv preprint. arXiv:1912.05784
Metadata
Title
Generalization of machine learning for problem reduction: a case study on travelling salesman problems
Authors
Yuan Sun
Andreas Ernst
Xiaodong Li
Jake Weiner
Publication date
01-09-2020
Publisher
Springer Berlin Heidelberg
Published in
OR Spectrum / Issue 3/2021
Print ISSN: 0171-6468
Electronic ISSN: 1436-6304
DOI
https://doi.org/10.1007/s00291-020-00604-x

Other articles of this Issue 3/2021

OR Spectrum 3/2021 Go to the issue