dc.description.abstract | Implementation of method mutations in genetic algorithms that are not coordinated
properly can make declining cause on an individual fitness value. To overcome these
problems do settings for use mutation methods that are expected to make the process of
finding a solution by genetic algorithms become more focused. In this study, the process
of finding a solution using genetic algorithm is then performed on some of the
individual with a minimum and maximum fitness value. Approach method used
mutations are mutations min-max, where individuals with a minimum fitness value
transferred in order to decrease and it is unlikely to survive, otherwise the individual
with maximum fitness value is transferred in order to increase the survival chances so it
can survive in the next process. In performance testing, a solution can be found quickly
and the number of generations needed by process can to be reduce, but the restrictions
on the number of generations is not always found a valid solution. Further testing with a
small number of individuals show better performance than the larger number of
individuals. | en_US |