基于二进制区分矩阵的属性约简方法及应用

The attribute reduction method based on binary discernibility matrix and applications

以粗糙集理论为基础,以属性重要度为启发式信息来指导约简过程,提出了一种改进的基于二进制区分矩阵的属性约简算法。以兰州市各区县主要经济林重点病虫害发生情况为例,使用该算法对影响病虫害发生的条件属性集进行约简,为决策者提供辅助决策支持,同时验证了该算法是有效可行的。

Abstract： The paper is based on rough set theory, the attribute significance as heuristic information to guide the reduction pro- cess, proposed an improved algorithm of attribute reduction based on the binary distinction matrix.And the main economic forest on plant diseases and insect pests of the districts and counties of Lanzhou city as an example, the use of the algorithm for reduction of condition attributes on effects of plant diseases and insect pests occurrence, provide auxiliary decision support to decision makers, and to verify that the algorithm is feasible and effective.