15- 100
80
20 : :0
y
Figure 4.2: Double-peak function
Genetic algorithms require that the parameter set of the problem be coded as a finite
length string. In addition, an objective function is necessary to measure how fit a current
solution is compared to the rest of a population. The search is guided using probability
and random choice. As the generations pass, strings associated with improved
performance will predominate and as the mating process progresses strings are combined
in new ways generating more sophisticated solutions [6].
Simple Genetic Algorithm
A simple genetic algorithm involves complex copying of strings and swapping of
partial strings. An initial population is chosen, usually at random. It is often necessary to
define a set of simple operations that take this initial population and generate successive
populations that improve over time.