LNS heuristic that tries to combine several feasible solutions.
- Author
- Timo Berthold
Crossover is a large neighborhood search improvement heuristic that is inspired by genetic algorithms and requires more than one feasible solution. For a set of feasible solutions, e.g., the three best found so far, it fixes variables that take identical values in all of them and solves a corresponding sub-SCIP. See also heur_mutation.h
Definition in file heur_crossover.h.