摘要： 针对遗传算法在全局优化问题中出现的早熟和收敛速度慢的问题，提出一种基于小生境遗传算法的粗糙集属性约简算法，采用基于淘汰相似结构机制的小生境技术，通过引入罚函数的方法调整个体的适应度，提高全局搜索能力。实验证明该算法是有效的，并能求解出信息系统中多组不同的最小约简，为决策支持和数据挖掘等提供更多信息。

关键词: 粗糙集, 小生境遗传算法, 属性约简, 数据挖掘

Abstract: To deal with the problems of prematurity and low convergence speed when Genetic Algorithm(GA) is used for global optimization, a rough set attribute reduction algorithm based on niche GA is proposed. Based on crowding mechanism, punishing function is adopted to adjust individual fitness. It can advance global capability. Experimental results show the algorithm is effective. It can find different reductions of attribute in the information system and provide more information for decision support and data mining.

Key words: rough set, niche Genetic Algorithm(GA), attribute reduction, data mining

中图分类号: 

• TP311