论文部分内容阅读
为了处理具有连续属性的决策系统,利用模糊理论与粗糙理论在处理不确定性问题方面的差异性,提出一种基于模糊-粗糙模型的逼近精度分类规则提取策略.首先利用模糊π函数对决策系统中的连续属性构造三个模糊参数进行模糊化,从而确定条件属性的模糊区域;再根据模糊相似关系构造模糊相似矩阵,然后基于模糊等价类划分的概念,提出了利用逼近精度近似度量的数据挖掘方法进行属性约减,最后应用实例说明如何在决策表中发现分类规则.实验结果表明此方法挖掘出的规则简练且合理可靠.
In order to deal with decision-making system with continuous attributes, this paper proposes a strategy based on fuzzy-rough model to extract approximation precision classification rules by using the difference between fuzzy theory and rough theory in dealing with uncertainty problems.Firstly, fuzzy decision- , The fuzzy attributes of three fuzzy parameters are fuzzified to determine the fuzzy region of the condition attributes. Secondly, the fuzzy similarity matrix is constructed based on the fuzzy similarities. Based on the concept of fuzzy equivalent classifications, the data using the approximate measure of approximation accuracy Mining method to attribute reduction, the last application example shows how to find the classification rules in the decision table.The experimental results show that the method of mining rules is concise and reasonable and reliable.