基于变精度粗糙集模型的属性约简的β值稳定区间讨论

Discussion of β Stable Interval of Attribute Reducts Based on Variable Precision Rough Set Theory

引入了基于变精度粗糙集模型的上、下分布属性约简方法的关于β值的稳定区间的概念,通过β值的稳定区间讨论,将β的取值区间[0,0.5]划分成为有限个稳定区间,当β值落在某个稳定区间内时,使得基于变精度模型的上、下分布约简不发生改变.不仅从理论上证明了β值稳定区间的个数是有限的,而且通过实例说明求解方法的简便和易于实现.为决策者提供了一种如何选择β值而获得上、下分布约简的参考.

Based on the variable precision rough set theory, the concept of β stable interval to upper or lower distribution reduction is introduced. When β is confined in some stable interval, the attribute reductions remain the same, so as to convert the infinite choices of β-value into finite ones. Not only the finity of the amount of β stable intervals is proved theoretically, but also the maneuverability and convenience of this method is interpreted by an example. So tbis paper provides the decision makers a reference on how to choose the β to gain different kinds of attribute reductions.