UNDERSTANDING AND USING GENETIC ALGORITHMS .1. CONCEPTS, PROPERTIES AND CONTEXT

Genetic algorithms are search algorithms founded upon the principles of natural evolution laid down by Darwin. They tum out to be competitive for a certain class of problems - complex large-scale problems, as a rule. Among the favorable properties of genetic algorithms are: efficiency, robustness and versatility. A less favorable property of genetic algorithms is the imprecision as a result of the noise used by the method. This tutorial consists of two parts which treat a variety of key issues concerning genetic algorithms. The first part emphasizes the principles underlying genetic algorithms, their search characteristics and the broader perspective in which they fit. This serves as a general, comprehensive introduction. Starting from the first part, the second part of this tutorial elaborates on practical issues such as representation, configuration and hybridization with other techniques. Thereby, some hands-on information is provided, so that common pitfalls can be avoided in using a methodology that exhibits its full power only when handled according to the principles it is based upon.