基于决策分类的决策表分解方法研究*

针对当前基于属性重要性的决策表属性集分解方法存在的不足，提出了一种新型的基于决策分类的决策表属性集分解方法。分析了近似分类质量和属性重要性与决策分类之间的关系，利用粗糙集理论，从提高子决策表中决策分类正确性的角度出发考虑条件属性与决策属性之间的关系，提出了决策表分解的条件属性选择量度并对决策表实施属性集分解。

Research on decomposition of decision table based on decision making

There are some defects in the attribute set decomposition of decision table based on the attribute significance. The paper introduced a novel feature decomposition of decision table approach based on decision making. It analyzed the diffe-rences between the quality of approximation and the attribute significance to decision making. With the rough set theory, the relationships of condition attributes and decision attributes were considered increasing the attribute classification rate in the decision table, and a new attribute selection criterion was proposed. Then decision table was decomposed with the attribute selection criterion.