Cooperative Interactive Cultural Algorithms Based On Dynamic Knowledge Alliance

Yi-Nan Guo,Shuguo Zhang,Jian Cheng,Yong Lin

DOI: https://doi.org/10.1007/978-3-642-23896-3_24

2011-01-01

Abstract:In cooperative interactive genetic algorithms, each user evaluates all individuals in every generation through human-machine interface, which makes users tired. So population size and generation are limited. That means nobody can evaluate all individuals in search space, which leads to the deviation between the users' best-liked individual and the optimal one by the evolution. In order to speed up the convergence, implicit knowledge denoting users' preference is extracted and utilized to induce the evolution. In the paper, users having similar preference are further divided into a group by K-means clustering method so as to share knowledge and exchange information each other. We call the group as knowledge alliance. The users included in a knowledge alliance vary dynamically while their preferences are changed. Taken a fashion evolutionary design system as example, simulation results show that the algorithm speeds up the convergence and decreases the number of individuals evaluated by users. This can effectively alleviate users' fatigue.