粗糙集知识约简中的蚁群优化算法研究  

Research on Knowledge Reduction in Rough Sets Based on Colony Algorithm Optimization

作　　者：赵建勋[1] 

机构地区：[1]中州大学,郑州450044

出　　处：《科技通报》2013年第8期53-55,共3页Bulletin of Science and Technology

摘　　要：蚁群算法具有全局优化和收敛速度上的优势,适合处理粗糙集的知识约简问题。因此,在求解决策表属性最小约简中,将属性的重要性作为启发式信息引入蚁群算法,并把相对核加入到初始信息素分布中。这样可以从整体上降低问题的规模,同时加快了系统的收敛速度。实验分析表明,本文提出的优化算法在实际应用过程中与传统算法相比,具有更好的综合性能。The ant colony has the advantage in global optimization and convergence speed.So it is suitable for dealing with the knowledge reduction in rough sets.Therefore,when looking for the solution for minimum reduction of decision table,we use the ant colony algorithm to take the importance if attribute as the heuristic information and add the relative core to distribution of original pheromones.The improved method can decrease the scale of problems in the whole and accelerate the convergence speed simultaneously. The experiments shows that the optimized algorithm has better comprehensive per- formance in actual implementation,comparing with the traditional algorithm ACO.

关 键 词：蚁群 粗糙集 约简 属性核 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]