论文部分内容阅读
遗传算法适合复杂问题的处理因此可用于属性约简的求解.目前利用遗传算法进行属性约简的主要不足是:适应度函数计算复杂,效率不高.尤其在处理大型决策表时,计算时间将大量聚集在适应度函数的计算上,从而导致算法性能下降.为了更快的计算适应度函数,在研究基于正区域的区分对象对集的基础上,设计了一种计算适应度函数的快速方法.利用启发信息设计了一种快速的属性约简遗传算法.通过实例分析和算法实验表明该算法能够高效求出决策表的属性约简并且适合处理大型决策表.
Genetic algorithm suitable for dealing with complex problems can be used to solve the attribute reduction.At present, the main disadvantages of using genetic algorithm to reduce the attribute are: the fitness function is computationally complex and inefficient, especially when dealing with large decision tables, computing time will be In order to calculate the fitness function more quickly, a fast method to calculate the fitness function is designed based on the study of the set of discriminant objects based on the positive region A rapid genetic reduction genetic algorithm is designed by using heuristic information.Experimental analysis and algorithm experiments show that this algorithm can efficiently obtain attribute reduction of decision table and is suitable for processing large decision table.