基于免疫算法的属性约简方法

为了求取决策表系统中属性的最小约简,本文提出了一种基于免疫算法的属性约简方法。该算法应用决策表的相对核来初始化种群,根据决策属性对条件属性的依赖度和抗体中条件属性的个数设计抗体的适应度函数,通过免疫记忆特性和抗体浓度的促进与抑制作用,保持了个体的多样性,提高了算法的全局搜索能力,避免陷入局部最优现象,从而求解出最小属性约简集合。实验结果表明,算法快速、有效,能得到较好的最小属性约简。

Attribute Reduction Approach Based on Immune Algorithm

In order to obtain the relatively minimal reduction of the attributes in a decision-making system, an attribute reduction algorithm is proposed based on immune algorithm. The core is joined initial population in the algorithm in order to accelerate capability. According to the dependability of decision attribute to the condition attribute and the condition attribute’s number of antibody, a new fitness function is defined. By the immune memory characteristics and the promoting and restraining function of antibody, it can maintain the individual’s diversity, and improve the global search ability of the algorithm, and avoid the local convergence, thus solves the minimal attribute reduction set. The experimental results show that the algorithm can find effectively and quickly the better minimal attribute reduction.