一种基于局部属性集的规则确定度

在传统的粗糙集中,利用确定度来评测决策规则的确定性,然而当多个决策规则拥有相同的确定度时,对给定的对象进行分类变得困难。基于局部属性集对决策规则确定性的影响,提出一种新的规则确定度，能充分反映决策规则的确定性在局部属性集上的差异。实例表明新的确定度对决策规则有较好的评测。

New certainty degree of decision rule based on local attribute sets

In traditional rough set, certainty degree is employed to evaluate the decision performance of a decision rule. However, it is difficult to choose one from some decision rules for classification of a given object when their certainty degrees are equal. A new certainty degree of decision rule based on local attribute sets was proposed. It sufficiently reflected the difference between certainty degrees of decision rules in local attribute sets. An example illustrates that new measure is effective to evaluate the decision performance of a decision rule.