Interactive genetic algorithms based on frequentpattern mining

In interactive genetic algorithms, user´s fatigue is the core problem. Aiming at this problem, implicit knowledge which presents the user´s variational preference is extracted to direct the evolution, at the same time, the speed of convergence is improved. Using frequent pattern algorithms to mine the implicit knowledge, frequent patterns toting the knowledge are extracted for every certain generations so that the knowledge could be update in time and premature convergence could be avoided. After being extracted, these frequent patterns are used to direct the evolution in the later generation. Taking the fashion evolutionary design system as example, the results of the simulation using the interactive genetic algorithms with frequent-pattern mining indicate that the algorithm can effectively alleviate users´ fatigue and improve the speed of convergence.