一种基于t相似分类的连续值域决策表的决策算法

A t Similar Classification Based Decision Algorithm for Decision Table with Continuous Condition Attributes

首先给出t相似度的定义,并引入t相似类的概念。随后,借助研究对象的t相似类定义了相似划分算法和广义决策集,并研究它们的性质,给出基于相似划分的类对应规则提取方法。接着,给出类对应规则中各条件类的区间表达,得到了面向连续值域决策表的规则提取算法。最后,结合实例说明了决策规则提取算法的实现过程。

First, the definition of t similarity degree between objects, t similarity class, similar partition operator, generalized decision set as well as their properties are discussed. Then, a decision algorithm for decision table with continuous attributes is provided. At last, the process of obtaining decision rules from continuous value decision table is explained by an example.