一种基于分类一致性的决策规则获取算法

Algorithm for acquisition of decision rules based on classification consistency rate

提出一种基于分类一致性的规则获取算法.它是一种例化方向的方法,即从空集开始,以条件属性子集的分类一致性来度量属性的重要性,逐步加入重要的属性,当选择的属性子集能够正确分类时,则获取到决策规则.算法中设计了一个规则约简过程,用来简化所获得的规则,增强规则的泛化能力.实验结果表明,所提出的算法获得的规则更为简洁和高效.

An algorithm for acquisition of decision rules, which uses the attribute importance measure based on classification consistency rate, is proposed. The algorithm uses a method by specialization, in which condition attributes are considered to measure the significance of the selected attributes set until the selected attributes set can make classification. A procedure for reduction of decision rules is also constructed. The procedure helps to get more precise rules which have more generalizing ability. The experiment and comparison show that the algorithm provides more precise and simple decision rules.