Hybridization Based on Problem Instance Reduction

  • Christian Blum
  • Günther R. Raidl
Part of the Artificial Intelligence: Foundations, Theory, and Algorithms book series (AIFTA)


This chapter presents an example of a hybrid metaheuristic for optimization based on the following general idea. General MIP solvers such as CPLEX and GUROBI are often very effective up to a certain, problem-specific instance size. When given a problem instance too large to be directly solved by a MIP solver, it might be possible to reduce the problem instance in a clever way such that the resulting reduced problem instance contains high-quality solutions—or even optimal solutions—to the original problem instance and such that the reduced problem instance can be effectively solved by the MIP solver. In this way, it would be possible to take profit from valuable Operations Research expertise that went into the development of the MIP solvers, even in the context of problem instances too large to be solved directly.


Problem Instance Solution Component Valid Solution Input String Instance Size 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer International Publishing Switzerland 2016

Authors and Affiliations

  • Christian Blum
    • 1
  • Günther R. Raidl
    • 2
  1. 1.Dept. of Computer Science and Artificial IntelligenceUniversity of the Basque CountrySan SebastianSpain
  2. 2.Algorithms and Data Structures GroupVienna University of TechnologyViennaAustria

Personalised recommendations