Document Open Access Logo

When Is Local Search Both Effective and Efficient?

Authors Artem Kaznatcheev , Sofia Vazquez Alferez

PDF

File

Thumbnail PDF
PDF
LIPIcs.STACS.2026.59.pdf
  • Filesize: 1.05 MB
  • 19 pages

Document Identifiers

Related Versions

Subject Classification

ACM Subject Classification
  • Theory of computation → Constraint and logic programming
Keywords
  • valued constraint satisfaction problem
  • local search
  • algorithm analysis
  • constraint graphs
  • pseudo-Boolean functions
  • parameterized complexity

Metrics

Abstract

Combinatorial optimization problems implicitly define fitness landscapes that combine the numeric structure of the "fitness" function to be maximized with the combinatorial structure of which assignments are "adjacent". Local search starts at an assignment in this landscape and successively moves assignments until no further improvement is possible among the adjacent assignments. Classic analyses of local search algorithms have focused more on the question of effectiveness ("did we find a good solution?") and often implicitly assumed that there are no doubts about their efficiency ("did we find it quickly?"). But there are many reasons to doubt the efficiency of local search. Even if we focus on fitness landscapes on the hypercube that are single peaked on every subcube (known as semismooth fitness landscapes, completely unimodal pseudo-Boolean functions, or acyclic unique sink orientations) where effectiveness is obvious, many local search algorithms are known to be inefficient. Since fitness landscapes are unwieldy exponentially large objects, we focus on their polynomial-sized representations by instances of valued constraint satisfaction problems (VCSP). We define a "direction" for valued constraints such that directed VCSPs generate semismooth fitness landscapes. We call directed VCSPs oriented if they do not have any pair of variables with arcs in both directions. Since recognizing if a VCSP-instance is directed or oriented is coNP-complete, we generalized oriented VCSPs as conditionally-smooth fitness landscapes where the structural property of "conditionally-smooth" is recognizable in polynomial time for a VCSP-instance. We prove that many popular local search algorithms like random ascent, simulated annealing, history-based rules, jumping rules, and the Kernighan-Lin heuristic are very efficient on conditionally-smooth landscapes. But conditionally-smooth landscapes are still expressive enough so that other well-regarded local search algorithms like steepest ascent and random facet require a super-polynomial number of steps to find the fitness peak.

Cite As Get BibTex

Artem Kaznatcheev and Sofia Vazquez Alferez. When Is Local Search Both Effective and Efficient?. In 43rd International Symposium on Theoretical Aspects of Computer Science (STACS 2026). Leibniz International Proceedings in Informatics (LIPIcs), Volume 364, pp. 59:1-59:19, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2026) https://doi.org/10.4230/LIPIcs.STACS.2026.59

Author Details

Artem Kaznatcheev
  • Department of Mathematics, and Department of Information and Computing Sciences, Utrecht University, The Netherlands
Sofia Vazquez Alferez
  • Department of Mathematics, and Department of Information and Computing Sciences, Utrecht University, The Netherlands

References

  1. Emile Aarts and Jan Karel Lenstra, editors. Local Search in Combinatorial Optimization. Princeton University Press, Princeton, NJ, USA, 2003. Google Scholar
  2. Yoshikazu Aoshima, David Avis, Theresa Deering, Yoshitake Matsumoto, and Sonoko Moriyama. On the existence of Hamiltonian paths for history based pivot rules on acyclic unique sink orientations of hypercubes. Discrete Applied Mathematics, 160(15):2104-2115, 2012. URL: https://doi.org/10.1016/J.DAM.2012.05.023.
  3. Taylan Batman. Arity of polynomials for equivalence classes of pseudo-Boolean functions. Bachelor’s thesis, Utrecht University, 2025. Google Scholar
  4. Umberto Bertelè and Francesco Brioschi. On non-serial dynamic programming. Journal of Combinatorial Theory, Series A, 14(2):137-148, 1973. URL: https://doi.org/10.1016/0097-3165(73)90016-2.
  5. Michaela Borzechowski. The complexity class polynomial local search (PLS) and PLS-complete problems. Bachelor’s thesis, Freie Universitat Berlin, 2016. Google Scholar
  6. Clément Carbonnel, Miguel Romero, and Stanislav Živný. The complexity of general-valued constraint satisfaction problems seen from the other side. SIAM Journal on Computing, 51(1):19-69, 2022. URL: https://doi.org/10.1137/19M1250121.
  7. David A Cohen, Martin C Cooper, Artem Kaznatcheev, and Mark Wallace. Steepest ascent can be exponential in bounded treewidth problems. Operations Research Letters, 48:217-224, 2020. URL: https://doi.org/10.1016/J.ORL.2020.02.010.
  8. William H. Cunningham. Theoretical properties of the network simplex method. Mathematics of Operations Research, 4(2):196-208, 1979. URL: https://doi.org/10.1287/MOOR.4.2.196.
  9. Marek Cygan, Fedor V Fomin, Łukasz Kowalik, Daniel Lokshtanov, Dániel Marx, Marcin Pilipczuk, Michał Pilipczuk, and Saket Saurabh. Parameterized algorithms. Springer International Publishing, 2015. URL: https://doi.org/10.1007/978-3-319-21275-3.
  10. Duc-Cuong Dang, Anton Eremeev, and Per Kristian Lehre. Non-elitist evolutionary algorithms excel in fitness landscapes with sparse deceptive regions and dense valleys. In Proceedings of the Genetic and Evolutionary Computation Conference, pages 1133-1141, 2021. URL: https://doi.org/10.1145/3449639.3459398.
  11. Duc-Cuong Dang and Per Kristian Lehre. The slo hierarchy of pseudo-boolean functions and runtime of evolutionary algorithms. In Proceedings of the Genetic and Evolutionary Computation Conference, pages 1551-1559, 2024. Google Scholar
  12. Benjamin Doerr and Frank Neumann, editors. Theory of Evolutionary Computation - Recent Developments in Discrete Optimization. Springer, 2020. Also available at URL: http://www.lix.polytechnique.fr/Labo/Benjamin.Doerr/doerr_neumann_book.html.
  13. MR Emamy-K. The worst case behavior of a greedy algorithm for a class of pseudo-boolean functions. Discrete Applied Mathematics, 23(3):285-287, 1989. URL: https://doi.org/10.1016/0166-218X(89)90018-8.
  14. Yahya Fathi and Craig Tovey. Affirmative action algorithms. Mathematical Programming, 34(3):292-301, 1986. URL: https://doi.org/10.1007/BF01582232.
  15. John Fearnley, Spencer Gordon, Ruta Mehta, and Rahul Savani. Unique end of potential line. Journal of Computer and System Sciences, 114:1-35, 2020. URL: https://doi.org/10.1016/J.JCSS.2020.05.007.
  16. Yuan Gao, Bernd Gärtner, and Jourdain Lamperski. A new combinatorial property of geometric unique sink orientations. arXiv preprint arXiv:2008.08992, 2022. Google Scholar
  17. Bernd Gärtner. The random-facet simplex algorithm on combinatorial cubes. Random Structures & Algorithms, 20(3):353-381, 2002. URL: https://doi.org/10.1002/RSA.10034.
  18. Teofilo F. Gonzalez, editor. Handbook of Approximation Algorithms and Metaheuristics: Methologies and Traditional Applications, volume 1. Chapman and Hall/CRC, New York, 2 edition, 2018. Google Scholar
  19. Mika Göös, Alexandros Hollender, Siddhartha Jain, Gilbert Maystre, William Pires, Robert Robere, and Ran Tao. Further collapses in tfnp. arXiv preprint arXiv:2202.07761, 2022. URL: https://arxiv.org/abs/2202.07761.
  20. Armin Haken and Michael Luby. Steepest descent can take exponential time for symmetric connection networks. Complex Syst., 2(2):191-196, April 1988. URL: http://www.complex-systems.com/abstracts/v02_i02_a03.html.
  21. Peter L Hammer, Bruno Simeone, Th M Liebling, and Dominique de Werra. From linear separability to unimodality: A hierarchy of pseudo-boolean functions. SIAM Journal on Discrete Mathematics, 1(2):174-184, 1988. URL: https://doi.org/10.1137/0401019.
  22. Thomas Dueholm Hansen and Uri Zwick. Random-edge is slower than random-facet on abstract cubes. Leibniz International Proceedings in Informatics, 55:51-1, 2016. URL: https://doi.org/10.4230/LIPIcs.ICALP.2016.51.
  23. Jun He and Yuren Zhou. Drift analysis with fitness levels for elitist evolutionary algorithms. Evolutionary Computation, 33(1):1-25, 2025. URL: https://doi.org/10.1162/EVCO_A_00349.
  24. Ronald A Howard. Dynamic programming and Markov Processes. John Wiley, 1960. Google Scholar
  25. Thomas Jansen and Ingo Wegener. Evolutionary algorithms-how to cope with plateaus of constant fitness and when to reject strings of the same fitness. IEEE Transactions on Evolutionary Computation, 5(6):589-599, 2002. URL: https://doi.org/10.1109/4235.974841.
  26. Johnny Jermias and Lindawati Gani. Integrating business strategy, organizational configurations and management accounting systems with business unit effectiveness: a fitness landscape approach. Management Accounting Research, 15(2):179-200, 2004. Google Scholar
  27. Robert G Jeroslow. The simplex algorithm with the pivot rule of maximizing criterion improvement. Discrete Mathematics, 4(4):367-377, 1973. URL: https://doi.org/10.1016/0012-365X(73)90171-4.
  28. David S. Johnson, Christos H. Papadimitriou, and Mihalis Yannakakis. How easy is local search? Journal of Computer and System Sciences, 37(1):79-100, 1988. URL: https://doi.org/10.1016/0022-0000(88)90046-3.
  29. Gil Kalai. Combinatorics with a geometric flavor. In Visions in Mathematics: GAFA 2000 Special volume, Part II, pages 742-791. Springer, 2000. Google Scholar
  30. Artem Kaznatcheev. Computational complexity as an ultimate constraint on evolution. Genetics, 212(1):245-265, 2019. Google Scholar
  31. Artem Kaznatcheev. Algorithmic Biology of Evolution and Ecology. PhD thesis, University of Oxford, 2020. Google Scholar
  32. Artem Kaznatcheev, David A Cohen, and Peter Jeavons. Representing fitness landscapes by valued constraints to understand the complexity of local search. Journal of Artificial Intelligence Research, 69:1077-1102, 2020. URL: https://doi.org/10.1613/JAIR.1.12156.
  33. Artem Kaznatcheev and Melle van Marle. Exponential steepest ascent from valued constraint graphs of pathwidth four. In 30th International Conference on Principles and Practice of Constraint Programming (CP 2024), pages 17:1-17:16, 2024. URL: https://doi.org/10.4230/LIPIcs.CP.2024.17.
  34. Artem Kaznatcheev and Sofia Vazquez Alferez. Greed is slow on sparse graphs of oriented valued constraints. In 31st International Conference on Principles and Practice of Constraint Programming (CP 2025), pages 44:1-44:13, 2025. URL: https://doi.org/10.4230/LIPIcs.CP.2025.18.
  35. Brian W Kernighan and Shen Lin. An efficient heuristic procedure for partitioning graphs. The Bell system technical journal, 49(2):291-307, 1970. URL: https://doi.org/10.1002/J.1538-7305.1970.TB01770.X.
  36. Victor Klee and George J Minty. How good is the simplex algorithm. Inequalities, 3(3):159-175, 1972. Google Scholar
  37. Mark W. Krentel. On finding and verifying locally optimal solutions. SIAM Journal on Computing, 19(4):742-749, 1990. URL: https://doi.org/10.1137/0219052.
  38. Daniel A Levinthal. Adaptation on rugged landscapes. Management Science, 43(7):934-950, 1997. Google Scholar
  39. J. Matousek and T. Szabo. RANDOM EDGE can be exponential on abstract cubes. Advances in Mathematics, 204:262-277, 2006. Google Scholar
  40. Jiří Matoušek. Lower bounds for a subexponential optimization algorithm. Random Structures & Algorithms, 5(4):591-607, 1994. URL: https://doi.org/10.1002/RSA.3240050408.
  41. Jiří Matoušek. The number of unique-sink orientations of the hypercube. Combinatorica, 26(1):91-99, 2006. URL: https://doi.org/10.1007/S00493-006-0007-0.
  42. Wil Michiels, Emile Aarts, and Jan Korst. Theoretical aspects of local search, volume 25. Springer, 2007. URL: https://doi.org/10.1007/978-3-540-35854-1.
  43. F.J. Poelwijk, D.J. Kiviet, D.M. Weinreich, and S.J. Tans. Empirical fitness landscapes reveal accessible evolutionary paths. Nature, 445:383-386, 2007. Google Scholar
  44. F.J. Poelwijk, T.-N. Sorin, D.J. Kiviet, and S.J. Tans. Reciprocal sign epistasis is a necessary condition for multi-peaked fitness landscapes. Journal of Theoretical Biology, 272:141-144, 2011. Google Scholar
  45. T Roughgarden. Computing equilibria: A computational complexity perspective. Economic Theory, 42:193-236, 2010. Google Scholar
  46. Alejandro A. Schäffer and Mihalis Yannakakis. Simple local search problems that are hard to solve. SIAM Journal on Computing, 20(1):56-87, 1991. URL: https://doi.org/10.1137/0220004.
  47. Ingo Schurr and Tibor Szabó. Jumping doesn’t help in abstract cubes. In International Conference on Integer Programming and Combinatorial Optimization, pages 225-235. Springer, 2005. Google Scholar
  48. Ingo A Schurr. Unique sink orientations of cubes. PhD thesis, ETH Zurich, 2004. Google Scholar
  49. Thomas Stützle. Local search algorithms for combinatorial problems: Analysis, Improvements, and New Applications. PhD thesis, TU Darmstadt, 1998. Google Scholar
  50. Tibor Szabó and Emo Welzl. Unique sink orientations of cubes. In Proceedings 42nd IEEE Symposium on Foundations of Computer Science, pages 547-555. IEEE, 2001. URL: https://doi.org/10.1109/SFCS.2001.959931.
  51. Melle van Marle. Complexity of greedy local search for constraint satisfaction. Master’s thesis, Utrecht University, 2025. Google Scholar
  52. Ingo Wegener. Methods for the analysis of evolutionary algorithms on pseudo-boolean functions. In Evolutionary Optimization, pages 349-369. Springer, 2002. Google Scholar
  53. D.M. Weinreich, R.A. Watson, and L. Chan. Sign epistasis and genetic constraint on evolutionary trajectories. Evolution, 59:1165-1174, 2005. Google Scholar
  54. S. Wright. The roles of mutation, inbreeding, crossbreeding, and selection in evolution. In Proc. of the 6th International Congress on Genetics, pages 355-366, 1932. Google Scholar
  55. Norman Zadeh. What is the worst case behavior of the simplex algorithm? In Polyhedral computation, volume 48 of CRM Proceedings and Lecture Notes, pages 131-143. American Mathematical Society, Providence, RI, 2009. Google Scholar
Any Issues?
X

Feedback on the Current Page

CAPTCHA

Thanks for your feedback!

Feedback submitted to Dagstuhl Publishing

Could not send message

Please try again later or send an E-mail
© 2023-2025 Schloss Dagstuhl – LZI GmbH About DROPS Imprint Privacy Contact