Abstract
State-of-the-art solvers for mixed integer programs (MIP) govern a variety of algorithmic components. Ideally, the solver adaptively learns to concentrate its computational budget on those components that perform well on a particular problem, especially if they are time consuming. We focus on three such algorithms, namely the classes of large neighborhood search and diving heuristics as well as Simplex pricing strategies. For each class we propose a selection strategy that is updated based on the observed runtime behavior, aiming to ultimately select only the best algorithms for a given instance. We review several common strategies for such a selection scenario under uncertainty, also known as Multi Armed Bandit Problem. In order to apply those bandit strategies, we carefully design reward functions to rank and compare each individual heuristic or pricing algorithm within its respective class. Finally, we discuss the computational benefits of using the proposed adaptive selection within the SCIP Optimization Suite on publicly available MIP instances.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Achterberg, T.: Conflict analysis in mixed integer programming. Discret. Optim. 4(1), 4–20 (2007). https://dx.doi.org/10.1016/j.disopt.2006.10.006
Achterberg, T., Koch, T., Martin, A.: Branching rules revisited. Oper. Res. Lett. 33(1), 42–54 (2005). https://dx.doi.org/10.1016/j.orl.2004.04.002
Berthold, T.: Heuristics of the branch-cut-and-price-framework SCIP. In: Kalcsics, J., Nickel, S. (eds.) Operations Research Proceedings 2007, pp. 31–36. Springer, Berlin (2008). https://dx.doi.org/10.1007/978-3-540-77903-2_5
Berthold, T.: Heuristic algorithms in global MINLP solvers, Ph.D. thesis, TU Berlin (2014)
Bubeck, S., Cesa-Bianchi, N.: Regret analysis of stochastic and nonstochastic multi-armed bandit problems. Found. Trends Mach. Learn. 5(1), 1–122 (2012). https://dx.doi.org/10.1561/2200000024
Computational Optimization Research at Lehigh Laboratory (CORAL): MIP instances. https://coral.ise.lehigh.edu/data-sets/mixed-integer-instances/
Forrest, J.J., Goldfarb, D.: Steepest-edge simplex algorithms for linear programming. Math. Program. 57(1), 341–374 (1992). https://dx.doi.org/10.1007/BF01581089
Gleixner, A., et al.: The SCIP Optimization Suite 5.0, Tech. Rep. 17–61, ZIB, Takustr. 7, 14195 Berlin (2017)
Harris, P.M.J.: Pivot selection methods of the devex lp code. Math. Program. 5(1), 1–28 (1973). https://dx.doi.org/10.1007/BF01580108
MIPLIB – the Mixed Integer Programming LIBrary. miplib.zib.de
Witzig, J., Berthold, T., Heinz, S.: Experiments with conflict analysis in mixed integer programming. In: Salvagnin, D., Lombardi, M. (eds.) Integration of AI and OR Techniques in Constraint Programming, pp. 211–220. Springer, Cham (2017). https://dx.doi.org/10.1007/978-3-319-59776-8_17
Acknowledgements
We thank Tobias Achterberg for useful comments and hints, especially with regard to Sect. 3.1. The work for this article has been partly conducted within the Research Campus MODAL funded by the German Federal Ministry of Education and Research (BMBF grant number 05M14ZAM).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Switzerland AG
About this paper
Cite this paper
Hendel, G., Miltenberger, M., Witzig, J. (2019). Adaptive Algorithmic Behavior for Solving Mixed Integer Programs Using Bandit Algorithms. In: Fortz, B., Labbé, M. (eds) Operations Research Proceedings 2018. Operations Research Proceedings. Springer, Cham. https://doi.org/10.1007/978-3-030-18500-8_64
Download citation
DOI: https://doi.org/10.1007/978-3-030-18500-8_64
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-18499-5
Online ISBN: 978-3-030-18500-8
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)