ABSTRACT
In any evolutionary computation application to an optimization problem, the human operator determines at least four aspects of the approach: representation, variation operators, method of selection, and objective function. It could be argued that the most crucial of these four is the objective function because it defines the purpose of the operator in quantitative terms. Improperly specifying the objective function can lead to generating the right answer to the wrong problem. However it should be clear that the selections made for each of these four aspects depend in part on the choices made for all the others. For example, the objective function cannot be specified in the absence of a problem representation. The choice for appropriate representation, however, cannot be made in the absence of anticipating the variation operators, the selection function, and the mathematical formulation of the problem to be solved. Thus, an iterative procedure for adjusting the representation and search and selection procedures in light of a specified objective function becomes necessary in many applications of evolutionary computation. This section focuses on selecting the representation for a problem, but it is important to remain cognizant of the interdependent nature of these operations within any evolutionary computation.
