广义特征矩阵及其应用初步

属性约简,即在保持知识库的分类或决策能力不变的条件下,删除其中不相关或不重要的属性,是Rough set理论的核心研究内容之一.决策表属性重要性度量方法是决定属性约简算法性能的重要启发式信息.合理的属性重要性度量方法,将有助于提高启发式搜索算法的效率和优化效果.针对基于分辨矩阵的属性重要性度量的缺陷,提出了广义特征矩阵概念,并在分析其性质的基础上,建立了一种新的基于广义特征矩阵的属性重要性分层度量方法,该方法不需要计算属性重要性的权值而直接给出重要性的排序,具有分辨能力强,度量准确的特点,对决策表的属性约简和知识荻取有重要应用价值.

Generalized eigenmatrix of decision table and its applications

The knowledge of reduction is one of the important issues in the research of rough set theory. Every heuristic information for the reduction of decision table must depends on a measure of the significance of attributes in the table, because it is the key heuristic information to determine which attribute is reducible in decision table. Based on the new definition of generalized eigenvector presented in this paper and analysis of its property, a new measure of the significance of attributes in decision table is given based on the generalized eigenmatrix. The new measure can discern the difference of significance of no-core attributes that cannot be discerned by the widely-used common measures. The new definition is very useful in the practice of data mining and knowledge reduction.