基于量子蚁群算法的粗糙集属性约简方法  

A Rough Set Attribute Reduction Method Based on the Quantum Ant Colony Algorithm

作　　者：袁浩[1] 

机构地区：[1]重庆邮电大学电子商务与现代物流实验室,重庆400065

出　　处：《计算机工程与科学》2010年第5期82-84,共3页Computer Engineering & Science

基　　金：重庆市科委自然科学基金资助项目(2009BB2288)

摘　　要：针对蚁群算法求取属性约简中存在的迭代次数多、收敛较慢甚至得不到最小约简的问题,提出了基于量子蚁群算法的粗糙集属性约简的方法。每只蚂蚁携带一组表示蚂蚁当前位置信息的量子比特;采用量子旋转门完成蚂蚁的移动;采用量子非门实现蚂蚁所在位置的变异。实验证明,该算法能快速有效地求解属性约简,同时又能找到许多次最小约简。可以很好地解决这一难题,它不仅能得到最小约简属性集,而且可以得到很多的约简属性集。As for the ant colony algorithm for attribute reduction which has the problems such as many iterationss, slow convergence obtaining no smallest reductions, this paper proposes the quantum ant colony for rough set attribute reduction. Each ant carries a group of ants which represent the current location information of quantum bits, uses quantum revolving doors to complete the ant movement, and uses quantum gates to realize the non-ant variation of the location. The experi ments show that the algorithm can quickly and efficiently solve attribute reduction, and find the smallest reduction. It not only can get the smallest reduction attribute set, but also get a lot of reduction attribute set.

关 键 词：属性约简 粗糙集 量子蚁群 蚁群算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]