Interactive genetic algorithms with interval fitness of evolutionary individuals

How to reflect the evaluation on evolutionary individuals assigned by human objectively is a difficult problem in interactive genetic algorithms (IGAs), and there is no effective method to solve it up to now. By applying interval analysis to evaluate evolutionary individuals in IGAs, an interactive genetic algorithm with interval fitness of evolutionary individuals is presented in this paper. The fitness of an evolutionary individual is an interval in the algorithm, and the width of the interval decreases gradually along with the evolution, which embodies that the cognition of human on evaluated objects is fuzzy and gradual. In addition, the probability that an evolutionary individual is dominant in tournament selection is based on interval dominance. The algorithm is applied to a fashion evolutionary design system, a typical optimization problem with an implicit index. It is shown from the results that the fitness of evolutionary individuals assigned in this paper is much objective.