Performance of Genetic and Evolutionary Algorithms on Tree Problems

Abstract

In the previous chapters, the presented theory about representations was mainly used for analysis and design of representations. The investigations into the properties of representations were based on theory and helped us to understand what happens when GEAs use a specific representation. However, in practice, GEA users are often less interested in theory about representations but want simple instruments for a fast and rough prediction of the expected performance of a representation. They have several representations at hand and want to know which representation they should choose for their problem. We do not want to leave them alone with their problems, but illustrate how they can advantageously use the proposed theory.