Hybrid Metaheuristics

Low-level relay hybrid: in which a given metaheuristic is embedded into a single-solution metaheuristic.

Low-level teamwork hybrid: a metaheuristic is embedded into a population-based metaheuristic.

High-level relay hybrid: self-contained metaheuristics are executed in sequence.

High-level teamwork hybrid: involves several self-contained algorithms performing a search in parallel and cooperating to find an optimum.

all hybrids us the same metaheuristic (Island model GA)

different metaheuristics are used

all algorithms search the whole search space

problems are decmposed into sub-problems, each having its own search space

solve the same target problem

combine algorithms which solve different problems (ex: Parallel tabu search solves QAP while GA makes a diversification task)

a set is applied one after the other, pipeline fashion

cooperation of optimization models where each agent carries out a search in the solution space and they cooperate.

a given function of a metaheuristic is replaced by a another metaheuristic function

the different metaheuristics are self-contained. There is not direct realtionship in the workings of the metaheuristics