Method of attributes reduction based on GA with memory function
一种具有记忆功能的遗传算法属性约简方法

Through the analysis of present algorism of attributes reduction for rough sets, focusing on the problems that the algorism of attributes reduction based on GA has, this paper proposed a new algorism of attributes reduction for rough sets based on a GA which had memory function. This algorithm made the GA have certain memory function by importing a flag bit, consequently, it could search in the two subgroups respectively. At the same time, it was only one subgroup that adopted the elitist model. All these could increase the probability of convergence and ensure correctness of the result. The experiment shows that the new algorithm is more effective than the algorithm of attributes reduction based on traditional GA.