ABSTRACT
Selection provides the driving force behind an evolutionary algorithm. Without it, the search would be no better than random. This section explores the pros and cons of a variety of different methods of performing selection. Selection methods differ in two main ways: the way they aim to distribute reproductive opportunities across members of the population, and the accuracy with which they achieve their aim. The accuracy may differ because of sampling noise inherent in some selection algorithms. There are also other differences that may be significant, such as time complexity and suitability for parallel processing. Crucially for some applications, they also differ in their ability to deal with evaluation noise.
