An introduction to Metaheuristic Techniques

We overview metaheuristics, applied to Combinatorial Optimization (CO) problems, and survey the most important classifications of metaheuristics, each of them being the result of a specific viewpoint. Furthermore, we describe the most important metaheuristics as they appear in the literature, along with variants and improvements. Although differences in how these algorithms tackle a problem are apparent, there are also several similarities in the strategies and concepts used by different metaheuristic techniques. The two most important concepts used in metaheuristics are intensification and diversification. They are in some way contrary and complementary to each other. In this work we compare metaheuristics in the way intensification and diversification are used, by introducing a framework that puts intensification and diversification mechanisms into relation with each other. We conclude outlining some strengths and weaknesses of the different approaches thus leading to the development of hybrid algorithms combining concepts originating from different metaheuristics.