Memetic Algorithms

Pablo Moscato

ABSTRACT

A memetic algorithm is a population-based approach for optimization that tries to synergetically use all possible information about the problem being addressed. This knowledge is generally, but not restricted to, some sort of previously-known heuristics or local search algorithms. We only describe here some of the most important applications of incomplete and stochastic memetic algorithm implementations when used as a metaheuristic. We particularly highlight their performance on  "classical" combinatorial optimization problems. These results can be regarded as benchmarks of the proposed methodology in comparison with other metaheuristics. We also give pointers to related work in the areas of scheduling, timetabling, computational biology, and molecular design.