Evolutionary Algorithms: A Tool for Addressing Problems which Humans Cannot Solve

Abstract:

Although computers have been applied in many areas, there are some applications which seem to be more difficult than others to computerise. Typically these are problems for which we do not have a complete understanding, such as computer vision or robot path planning. Traditional development methods cannot account for a poor analysis of a problem and therefore fail to deliver successful systems for ill-defined problems. Three case studies are presented to demonstrate the application of genetic algorithms and genetic programming to demonstrate how these evolutionary techniques can be applied to ill-defined problems, thus diminishing the need for humans to apply themselves to dangerous or mundane tasks.