Understanding evolutionary computing: a hands on approach

Evolutionary computing is the study of robust search algorithms based on the principles of evolution. An evolutionary algorithm (EA) searches a problem space in order to find regions containing good solutions. Typically EA users judge the quality of their algorithms by the quality of the solutions found. This approach ignores the behavior of the search algorithm and concentrates solely on the outcome. As a result, the user is unaware of their algorithm's actions or how the solutions were discovered. This paper describes how search space visualizations can be used to facilitate the user's understanding of evolutionary computing. A set of examples are presented showing how the user can take a "hands on" approach to explore the behavior of their algorithms and interact with the evolutionary search process.