交互式遗传算法基于用户认知不确定性的定向变异

Directional mutation based on user's uncertain cognitive in interactive genetic algorithms

利用用户认知的不确定性设计定向变异算子.首先,采用主成分分析法辨识用户认知的不确定性;然后,给出用户认知不确定性的区间表示与更新策略;最后,将用户认知的不确定性指导定向变异算子,包括:选择待变异的进化个体,确定变异位置,以及变异方法等.将所提方法应用于人眼图形优化,实验结果验证了该方法的优越性.

A directional mutation operator is designed by using the user＇s uncertain cognitive.The main ingredient analysis is used to identify the user＇s uncertain cognitive.An interval-based expression and an update strategy of the user＇s uncertain cognitive are given.Finally,the user＇s uncertain cognitive is to guide a directional mutation operator,including the choice of candidate individuals to be mutated,the determination of positions to mutate,as well as the strategy to mutate.The proposed algorithm is applied to eyes optimization.The experimental results show the advantageous of the algorithm.