论文部分内容阅读
决策形式背景的最小约简可以使规则提取更加简便,也可以使所获取的规则更加紧凑,从而有利于数据的决策分析.对于如何快速求得决策形式背景的一个最小约简,已有一些启发式方法在这方面做了有益的尝试.然而,启发式思想求解最小约简遇到某些特殊的数据集会出现失效的现象.在决策形式背景中引入决策规则支持元与支持度,讨论了协调集与约简的等价判定定理,在此基础上提出基于遗传算法的决策形式背景的属性约简方法.数值实例分析表明,新约简方法能够在一定程度上弥补启发式算法存在失效现象的不足,从而有利于提高决策分析的效率.
Minimal reduction in decision-making formal context can make rule extraction easier, and also make the rules obtained more compact, which is beneficial to the data decision analysis.However, there is some enlightenment on how to get a minimal reduction of the decision-making context quickly However, heuristic thinking solves the minimal reduction and encounters the failure of some special data sets.In the context of decision-making, the introduction of decision-making rules support meta and support, discusses the coordination Set equivalence and reduction equivalence decision theorem, based on which the attribute reduction method based on genetic algorithm decision-making formal context is proposed.Numerical example analysis shows that the new reduction method can make up for the heuristic algorithm to some extent, the failure phenomenon Inadequate, which will help improve the efficiency of decision analysis.